Forums :: Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives :: Books
Support This Site: Contributors :: Advertise


It is currently Aug Thu 21, 2014 7:03 pm


All times are UTC [ DST ]



Post New Topic Post Reply  [ 27 posts ]  Go to page 1, 2  Next
Author Message
 Post subject: Some Fender amp info
PostPosted: Oct Fri 17, 2008 2:40 pm 
Member
User avatar

Joined: Jun Thu 29, 2006 2:37 pm
Posts: 475
Location: Norwalk, CT
Hi All,

After about 30 years of messing with Fender amps I have settled on my setup philosophy. I just learned something new that Fender people would like to know.

It came about because of a 50 watt Bassman Ten that I aquired. This amp was so up-front sounding that I had to look into why.

I tested my blackface Bandmaster and silverface Bandmaster Reverb Amp against it. I scoped them out together for frequency response. After I setup the tone circuits to match the Bassman, the 3 amps had the same exact response in steady-state spot checks. There was some phase difference at high frequencies, and this is a very audible factor in a guitar amp. Even though frequency response was the same for all amps, the Bassman sounded so much different and worked so much better in a band situation.

Lots of head-scratching and experimenting...

I should note that first I had to tame the parasitics in both of my silverface amps - that was easy.

Finally I setup the silver Bandmaster power supply to match the Bassman - that is I eliminated the Pi filter choke and replaced it with a 2Kohm 10 watt wirewound resistor. I could not believe my ears. The amp became so much more clear and alive.

I am now in favor of the CBS phase inverter values as they help prevent farting and barking that occurs from grid-blocking and diode-clamping in the outputs at peak signal. This, I read on the Web, is due to time-constant changes. Besides- I like the sound of less gain in the PI stage. I also went with SS rectification and cut the value of the coupling caps to the outputs from .1uf to .047uf. The tone circuit is like the Bassman Ten. I also used the .022uf with 1mohm to chassis at the input of the PI to emulate the Bassman Ten Amp, and used a .01uf at the PI input. I disconnected the "Normal" channel mixing resistor, and jumped the mixing resistor in the "Vibrato" channel. This makes the PI input work much better. I also put a 500pf snubber across the plate resistor of the first stage. This gets rid of that spiky ringing (as seen on scope) that is typical of Fenders when the treble and volume controls are up near full.

It became up-front sounding. I giged with it last night and (after 30 years of trying) I finally got my sound. The bottom-end is always perfect now. It never gets muddy when cranked as is the typical scenario with a Fender. The beauty sound is there. The amp is totaly touch-responsive. It will be clean or get the most beautiful honk/grit when I dig-in. Mids are strong, balanced and clear, treble is sparkly and not piercing. No graininess or hash. I'm in love with this amp. 50 watts into two -12" Jensen reissue C12Q in a Carvin open-back cabinet. The amp and speaker breakup match perfectly.

So that's it. The main point being that I finally did the experiment of getting the choke out of circuit and the amp now became textured, tight and strong. Why didn't I try that in 1983 or something? The CBS Bassman Ten circuit put my attention there.

So let's score one for the rocket scientist that went to work for Fender and promptly ruined thier reputation with those other changes. They did good on the 50 non-ultralinear Bassman Ten.

Thought I would mention it here, as there are some Fender guys, and I have seen no mention of this by Gerald Weber or the other Fender gurus, or anyone else.

Unbelievable that it should do so much when the choke is replaced with a resistor - that was the major breakthrough to the sound.

Why? We can speculate...

Also remember that the screen grids are at that exact point on the supply.

I am estatic about this (can you tell?) - our guitar sound is the thing.

Try it - you'll like it. You won't need NOS/boutique tubes with this setup.

"God gave rock and roll to you" Rod Argent.

Denis


Top
 Profile  
 
 Post subject:
PostPosted: Oct Fri 17, 2008 7:32 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 11316
Location: Warner Robins, GA
By using the resistor youi may have lowered screen voltage a little provided the choke does not have the same DC resistance as the resistor. That may have put the output tubes in a more linear operating mode.


Top
 Profile  
 
 Post subject:
PostPosted: Oct Fri 17, 2008 7:33 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3296
Location: WI 54812
Denis-

Interesting stuff.

Would explain why a lot of the older amps like the 6E5 Deluxe are so popular, also the good sounding Princeton’s. Perhaps that choke is filtering out harmonic distortion like they originally wanted it to? I also think the resistor would allow for more of a compression envelope in the power supply(sag). Of course, the reason they took those chokes out was to save money on the cheaper amps, and also later with the CBS circuit mods. IMO, in general, stock Fender amps form the 60-70's have too much bass response with anything other than a Tele. I had to tame the bass down on my Princeton Reverb clone I did. Otherwise my Hamer Special sounded real boomy through it. Also, I think a 12at7 works much better than a 12ax7 in the PI circuit. It allows the power tubes grids to draw more current before the PI tube clips hard. Gets the thin fuzzy sound out of it.

Glad you finally got -your- sound nailed down. I personally feel a amp needs to be tuned to the guitar and the person playing through it. Everything is a compromise with electronic design.

Kevin


Top
 Profile  
 
 Post subject:
PostPosted: Oct Fri 17, 2008 9:33 pm 
Member
User avatar

Joined: Jun Thu 29, 2006 2:37 pm
Posts: 475
Location: Norwalk, CT
Tube Radio wrote:
By using the resistor youi may have lowered screen voltage a little provided the choke does not have the same DC resistance as the resistor. That may have put the output tubes in a more linear operating mode.


Right TR - the screen supply is now 34v below the B+. With the choke it was a 2 or 3 volt drop. The whole spectrum became free of confusion and garbling.

Interesting Kevin - I was never a tweed guy, but that is mostly by coincidence and also I like the front-facing controls. Plus a little reverb is nice. Think I'll look over those tweed schematics.


Top
 Profile  
 
 Post subject:
PostPosted: Oct Fri 17, 2008 10:50 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3296
Location: WI 54812
If its the lower screen supply voltage making the difference in sound, why couldn't one adjust the size of the screen grid resistors upward and leave the choke alone? I don't think droping the screen voltage lower is making the output -more- linear. The screen isn't really getting any feedback from the plate circuit in this condition is it? It would be fun to set up an adjustable screen supply voltage and see how that affects the tone and overload while you play.

Kevin


Top
 Profile  
 
 Post subject:
PostPosted: Oct Sat 18, 2008 2:06 am 
Moderator

Joined: Jan Thu 01, 1970 1:00 am
Posts: 8140
Location: Latham NY
What was the the DC resistance of the choke? You replaced it with a 2K resistor which I would think is a lot higher than the choke would have been. Did you change the value of the filter caps?


Top
 Profile  
 
 Post subject:
PostPosted: Oct Sat 18, 2008 2:21 pm 
Member
User avatar

Joined: Jun Thu 29, 2006 2:37 pm
Posts: 475
Location: Norwalk, CT
Lou deGonzague wrote:
What was the the DC resistance of the choke? You replaced it with a 2K resistor which I would think is a lot higher than the choke would have been. Did you change the value of the filter caps?


I don't remember the last time I measured that Lou, but I will.

I think the screen voltage/linearity idea holds water, but there is something else maybe.

These amps get some tuning (as you may be suggesting) by the choice of decoupling. Especially the final filter to the front-end. One can tune the softness/stiffness here to match the speakers being used. I did not alter the decoupling values, but I am thinking that the use of a resistor in the Pi filter changes the phasing of the power supply feedback interaction which is a big part of the character.

Phasing is very audible. It's the difference between an amp that sounds good and one that produces a "watery" tone that washes over you and makes you smile.

I think that maybe the amp is now more phase coherant in it's PS feedback dynamics, and that is why it lost the garbling and smearing. (it did sound good before - but now it's on a whole 'nother level).

So when the audiophools talk about the power supply being important they may take it too far sometimes, but they are not wrong.

So for now I will think about phasing and linearity.

So in respect to what Kevin said about just changing the screen resistor values - I think you are right that this will not do the trick all by itself. I have reason to think that the removal of a highly inductive component is the thing.

What do you think?

Saw The Venture lastnight at B.B. Kings. A blonde Fender something with spaghetti logo for Nokie Edwards (lead player) - very clear. Don Wilson had an outboard Fender Reverb unit that was up so high you could here the sproing of the tank springs slapping around. In the next room was a blues guy who told me he carried his Twin Amp 4 blocks to get to the gig - I am suspicious of his mental health. We had fun.

Denis


Top
 Profile  
 
 Post subject:
PostPosted: Oct Sat 18, 2008 3:13 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3296
Location: WI 54812
dsbk wrote:

Phasing is very audible. It's the difference between an amp that sounds good and one that produces a "watery" tone that washes over you and makes you smile.

I think that maybe the amp is now more phase coherant in it's PS feedback dynamics, and that is why it lost the garbling and smearing. (it did sound good before - but now it's on a whole 'nother level).


I'm having trouble wrapping my mind around the possibility of the power supply influencing phasing at anything other than very low audio frequencies. But you're obviously hearing something. In other words the electrolytic PS caps are not effective at decoupling all the audio frequency ripple(feedback), but the choke was?

I think you owe us a sound clip so we can witness the watery sound of this amp :)

Kevin


Top
 Profile  
 
 Post subject:
PostPosted: Oct Sat 18, 2008 3:30 pm 
Member
User avatar

Joined: Jun Thu 29, 2006 2:37 pm
Posts: 475
Location: Norwalk, CT
Kevin Kuehn wrote:
dsbk wrote:

Phasing is very audible. It's the difference between an amp that sounds good and one that produces a "watery" tone that washes over you and makes you smile.

I think that maybe the amp is now more phase coherant in it's PS feedback dynamics, and that is why it lost the garbling and smearing. (it did sound good before - but now it's on a whole 'nother level).


I'm having trouble wrapping my mind around the possibility of the power supply influencing phasing at anything other than very low frequencies. But you're obviously hearing something. In other words the electrolytic PS caps are not effective at decoupling all the audio frequency ripple(feedback), but the choke was?

I think you owe us a sound clip so we can witness this watery sound of this amp :)

Kevin


Not the ripple Kev, but the actual audio that gets on the supply. The result is part of what the gurus call the envelope of the amp - as you may know. Those are not ripple filters, just decouplers to keep some of the audio out of the supply. Audio in the supply modulates the whole amp. You know all this, but I put it here for record and discussion.

I think the audio in the PS feedback is what made the change. Maybe it becomes less smeared in time.

The change was not in frequency response, but in the way the same frequencies are presented. Clarity - coherance. It is an apparant change in frequency response, or maybe a real one in terms of the whole sound dynamicly as it comes out of the amp with different phase reinforcements and cancellations. Like when you adjust a tape head by putting the channels together in mono and finding the best phase with the least swarshing.

If we decouple completely - like a hi-fi guy would do the amp becomes more sterile sounding for guitar. It becomes stiff, which may be good for your stereo.

I think the thing here is phasing in the audio that is on the PS, which feeds back to every stage. Maybe the choke produces other components as well. This may sound like smoke and mirrors to some folks, but it is a very plain fact-of-life for us. Did you ever overbuild a Fender supply and find that the "beauty" of the amp was lost? It moves into Marshall territory.

Yes, I see that the old Super, Deluxe, Twin and other amps used the 10K resistor. So the CBS engineers actually reverted back to that when they did the Bassman Ten.

How to do a sound clip? I mean where does one put it? I can make a wav file or MP3 tonight.

Denis

_________________
"Racing is life... everything before and after is just waiting." Steve McQueen in 'Le Mans'


Top
 Profile  
 
 Post subject:
PostPosted: Oct Sat 18, 2008 5:24 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3296
Location: WI 54812
I guess I always envisioned an amps envelope as the overall lowering/rising (modulation) of the PS voltages, as when one hits a power cord. I would have to believe most of that sag occurs as a result of the power tubes drawing more current. I know a rectifier tube can have a big influence on the PS envelope dependant on its internal impedance to the load. That PS sag at the output end would in-turn lower the voltage to the previous pre-amp stages. What I don’t get is how there can be much actual –audio signal- riding on the PS because of this. Even though the audio may modulate the average PS voltages, it’s beyond me how this would provide actual audio feedback to the previous stages. In theory, the ps decoupling caps should not allow any audio from the output stage to be transmitted to the input stages. Have you scoped an audio signal on the various PS stages with a strong test signal on the amp input?

Not trying to be a stick in the mud, just would like to get a better understanding of what’s causing whatever it is you’re experiencing.

I could temporarily host a sound clip on my web space if its not huge. It will take a while for me to load it, as I’m on dial up here. What would be really interesting is if you could do a before and after choke clip. I’m wondering if most people will actually pick up on the difference. If you want to try something like this I would pm you my email address. No guarantees it will work.

Kevin


Top
 Profile  
 
 Post subject:
PostPosted: Oct Sat 18, 2008 5:33 pm 
Member
User avatar

Joined: Jun Thu 29, 2006 2:37 pm
Posts: 475
Location: Norwalk, CT
OK, to do it right will take some preparation. Probably not over the weekend.

On some amps the audio just gets everywhere. Sag would be a little different from the decoupling effect.

I like sag and beauty on amps that will be played in their own atmosphere. Lately I have been hosting an open mic at a local cafe'. We have a huge stage and the PA sounds like a hi-fi. I'm in the house band. We get plenty of jammers and guitar monsters. I get a chance to hear all my amps and test again after I make changes. What I am seeing is that a more stiff and less bloomy/beautiful sound cuts better and becomes fluid in the presence of full instumentation. My beauty amps got lost in the sauce and muddy.

The dynamics of acoustics and physco-acoustics is just a wonder.

Denis

_________________
"Racing is life... everything before and after is just waiting." Steve McQueen in 'Le Mans'


Top
 Profile  
 
 Post subject:
PostPosted: Oct Sun 19, 2008 6:45 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3296
Location: WI 54812
Was thinking about this some more…

I have a hunch what you are experiencing is mostly caused by lowering the Vs (screen grid voltage) on the output tubes. As was mentioned earlier- with a choke Pi filter, Vs is much closer to Vp (plate voltage). I doubt early Fender ever intended these amps to be pushed into class AB2 operation. With a choke filter, when the output tubes draw grid current, Vp likely falls below Vs causing screen grid current to flow. I think that is what contributes to the ratty harsh sound at clipping and a smeared sound in general. Look at any tube manual and they show Vs quite a bit lower than Vp, at least for anything other than class A operation. The addition of the 2k resistor drops Vs and limits screen current under high loads. This in turn allows the output tubes to compress some. The perception is a tighter more up front sound. It acts as a sort of limited automatic volume control on the output stage. The trade off is that as you decrease Vs you also loose power (dynamics) in the output stage. So, it becomes a balancing act.

Sorry, but I’m just not willing to accept that significant audio signal (feedback) from the power section is getting past those 20uf-decoupling capacitors and into the front sections. :)

Kevin


Top
 Profile  
 
 Post subject:
PostPosted: Oct Sun 19, 2008 3:54 pm 
Member
User avatar

Joined: Jun Thu 29, 2006 2:37 pm
Posts: 475
Location: Norwalk, CT
OK Kevin - this is good. We diverge here in our idea of what is happening when the choke is replaced with the 10 Kohm resistor.

You are leaning toward thinking that the screen voltage is affecting dynamics in the output section. Linearity and such.

I feel that there is a phasing phenomenon going on in the PS that becomes corrected.


I agree the screen volts change may be part of the help dynamically, but not with the clearing of the tone (which I say again was sudden with the choke/resistor swap, major and very, very obvious to the point where you would think it were a different amp or speakers). The swap was done after all other changes, with exception of 1 cathode cap (done later). I was concentrating on tones and was out of ideas when I tried the magic swap.

Let's find out.

This requires an experiment. The thing is that quantifying is done with the ears. But I did see a high-frequency phase difference on the scope when the two amps were run concurrently. First I had set the controls on one amp to the way I use them, then ran in a 1Khz square wave for about a 1 watt output, then matched the 2nd. amp up to superimpose the waveforms exactly. Then I did spot frequency test. The amp with the larger cathode and coupling caps went down a little lower in bass. I have since trimmed that out some, on the cathode end of things by putting in the 5uf K cap on the Bandmaster Reverb Amp to match the Bassman Ten. The Bassman retains the .1uf coupling caps while Bandmaster has the .047uf in it.

Both PI circuits are the CBS low resistance style, and both have a .01 to PI input. I used a 56K on the inboard side of the Bandmaster plate to even-out the clipping. The Bassman clipped evenly with both plates using 47K, so I think the output transformer is wound differently. The bias resistors are all 100Kohm.

Both amps have a .001 directly across the PI plates for control of parasitic oscillation that they both exibited when the .002 tone-killers were removed from the output grids. This is the way I found that I like to do it.

The Bassman has 50 then 20 then 20 then 20uf caps.
The Bandmaster has 40 then 20 then 20 then 20uf caps.

Both amps use a 2.2k and then 10k decoupling resistor for PS pick-off points C and then D.

Both amps use SS rectification with a B+ of almost 500v when running at idle. Output tubes are biased at around 42 Ma each.

The Bandmaster has the additional voltage amp after reverb recovery. I decreased the coupling from .1 down to .022uf for the output of that extra stage.

Both amps are setup for a single channel only. I could use the NORMAL channel in the Bandmaster to make the test more alike. At this time it is disconnected and the VIBRATO channel mixing resistor is bypassed.

The Bandmaster produces just a few more watts (52watt rms) than the Bassman at mild clipping.

I think that a recorded listening test and then to repeat the spot frequency test would be good.

Maybe to scope for audio on the PS would be good.

I don't have a spectrum analyzer.

Jim Marshall (owner of THD amps) would be interested in this experiment. I spoke with him years ago and found out he loves to tinker like this. I'll bet he has an opinion on it. Will try to give him a call.

Have you got a proceedure you would like to see done?

Denis

_________________
"Racing is life... everything before and after is just waiting." Steve McQueen in 'Le Mans'


Top
 Profile  
 
 Post subject:
PostPosted: Oct Sun 19, 2008 6:11 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3296
Location: WI 54812
dsbk wrote:
I feel that there is a phasing phenomenon going on in the PS that becomes corrected.

There very well may be phase related changes going on.

Don't underestimate the side effects of changing Vs. Remember, this is changing the operational point on the tubes load line curve. This could explain the phase differences you are seeing/hearing. My point is- that I tend to think any phase corrections are occurring in the output transformer to input of the PI feedback loop, rather than from the output stages of the power supply feeding back to the front stages.
dsbk wrote:
Maybe to scope for audio on the PS would be good.

You will see some audio on output transformer center tap end, as that's where the most current is being drawn. But, this is the exact purpose of the PS decoupling caps, to smooth any 120hz ripple, and also isolate and decouple feedback between the audio amplification stages. The various filtering stages progressively reduce PS ripple, more toward the front end. This is important because the front end is the area of most gain. If the caps are large enough to filter out 120 Hz ripple, they're also not going to pass much in the order of audio signals between stages.

As another experiment, try inserting your test signal directly to the PI input. Then take note of how altering Vs effects frequency response and overload. Try this with both the Pi choke and the 2k resistor. This will eliminate any tonal variations from the front end. It would be pretty easy to utilize a DPST switch between the two filter types.

It's unfortunate that you're not getting more input from our members with a engineering background. I'm really not very good at explaining the theory behind varying Vs and how it interacts with feedback and frequency response. I do know it can have a very profound effect on how the power tubes overload.

I'll try to butt out and let someone else have a turn. :)

Looking forward to the outcome of your further experimentation. This is a very fun subject to me.

Kevin


Top
 Profile  
 
 Post subject:
PostPosted: Oct Sun 19, 2008 7:08 pm 
Member
User avatar

Joined: Jun Thu 29, 2006 2:37 pm
Posts: 475
Location: Norwalk, CT
Yeah Kev, I would like to hear what the engineers have to say also. I don't know that this subject is very interesting for them. If you are not a player then it would probably seem like hocus-pocus, while for us the difference is huge.
Denis

_________________
"Racing is life... everything before and after is just waiting." Steve McQueen in 'Le Mans'


Top
 Profile  
 
 Post subject:
PostPosted: Oct Sun 19, 2008 10:56 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 1953
Location: Hinsdale, IL, USA
Kevin and Deni,

Interesting discussion. I have zero guitar amp experience, but I'll put my 2 cents in here.

I did some simulations with PP 6L6GCs I used a purely resistive load of 5k, +B 400 V, plate current 42 mA and I varied the screen voltage from 250 V up to 400 V. The fixed grid bias was adjusted to alway maintain 42 mA per tube with no signal. I used a peak grid signal of 20 V.

Basically here's what I found. (Granted I did not build the circuit to verify this. I have a 6V6 PP amp I might play with to see if it behaves the same as my simulation.)

Starting with the Vs at 400 V the output power was 33 W and THD 2%. As I brought the Vs down the power increased and the THD decreased. The minimum THD (1.45%) occurred with the Vs equal to 320 V the power was 39 W at this point.

This assumes perfectly matched tubes and transfomer. The even harmonics are all zero and the THD is calculated up to the 9th harmonic.

FWIW

_________________
Chuck D. KB9UMF


Top
 Profile  
 
 Post subject:
PostPosted: Oct Sun 19, 2008 11:21 pm 
Member
User avatar

Joined: Jun Thu 29, 2006 2:37 pm
Posts: 475
Location: Norwalk, CT
CD, you always bring something good in. :) I did not know that a power change could happen like that.

_________________
"Racing is life... everything before and after is just waiting." Steve McQueen in 'Le Mans'


Top
 Profile  
 
 Post subject:
PostPosted: Oct Sun 19, 2008 11:24 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3296
Location: WI 54812
Cdoose wrote:
I did some simulations with PP 6L6GCs I used a purely resistive load of 5k, +B 400 V, plate current 42 mA and I varied the screen voltage from 250 V up to 400 V. The fixed grid bias was adjusted to alway maintain 42 mA per tube with no signal. I used a peak grid signal of 20 V.


What are you using to create the simulation?


Cdoose wrote:
Starting with the Vs at 400 V the output power was 33 W and THD 2%. As I brought the Vs down the power increased and the THD decreased. The minimum THD (1.45%) occurred with the Vs equal to 320 V the power was 39 W at this point.
FWIW
This is really interesting. So what happenes when you bring Vs below 320v, did power drop and distortion increase? In other words there is a happy medium point for Vs, in order to get the least distortion and most power?

[edit] Interestingly RC19 recommends a Vs of 270 with a Vp of 360.

Kevin


Top
 Profile  
 
 Post subject:
PostPosted: Oct Mon 20, 2008 12:20 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 1953
Location: Hinsdale, IL, USA
I'm using an MS Excel spread sheet simulation that I designed. I trust it very much for triodes, but I haven't confirmed it's accuracy for pentodes or beam power tubes yet. I use actual measured plate curves and fit these curves with equations to generate the tube characteristics. Way too involved to explain in a few words.

The sweet spot for the Vs in this simulation occurred at 320 V below that the distortion really gets bad in a hurry. That's because the maximum plate current is below the knee of the plate curve for the load line I was using (5k plate to plate), and the grid bias is approaches the peak grid signal. This will cause clipping.

I will try and confirm this with some real measurements and let you know.

_________________
Chuck D. KB9UMF


Top
 Profile  
 
 Post subject:
PostPosted: Oct Mon 20, 2008 2:27 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3296
Location: WI 54812
This is an interesting read if you’re into this weird guitar amp responsiveness.
http://www.freepatentsonline.com/4713624.html
Scroll down to description. Patent holder is Randall Smith of Mesa Boogie fame.

Kevin


Top
 Profile  
 
Post New Topic Post Reply  [ 27 posts ]  Go to page 1, 2  Next

All times are UTC [ DST ]


Who is online

Users browsing this forum: No registered users and 4 guests



Search for:
Jump to:  



















Privacy Policy :: Powered by phpBB