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


It is currently Sep Wed 19, 2018 1:40 pm


All times are UTC [ DST ]





Post New Topic Post Reply  [ 21 posts ]  Go to page 1, 2  Next
Author Message
 Post subject: Buffer amp
PostPosted: Mar Tue 06, 2018 2:22 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19018
Location: Warner Robins, GA
I have a 70Hz two way 24 db/octave active crossover that I currently use in my second stereo system to drive the main amp and a bucket sub.

Attachment:
Power supply.jpg
Power supply.jpg [ 49.19 KiB | Viewed 907 times ]


Attachment:
HP LP filter LP sum.jpg
HP LP filter LP sum.jpg [ 91.33 KiB | Viewed 907 times ]


I am looking at getting a Dayton Audio plate amp with built in DSP which also has an adjustible high pass crossover thusfore making this crossover unnecesary.

Now due to the input impedance of the plate amp I will need to buffer the output of the McIntosh C24 preamp I use plus I will need to buffer the record out jacks of the preamp.

My thought is to modify this crossover only using one quad OP-AMP to serve as a buffer for the preamp output and record output. Is that a good idea or should I use two of the OP-AMPs using only two sections of each?

The hard part is I built this on the perfboard with solder pads on one side so it won't be easy to remove components.

I would like to get rid of the circuit alltogether if possible.

Is there any way to add a JFET in the McIntosh preamp to buffer the record out and add a TL-082 to the plate amp provided there is a source of + and - 15-18Vdc on board the plate amp?

The only concern about the JFET is the fact that the McIntosh power transformr is rated at 20mA on the secondary, although in previous experience with a JFET buffer I used for a stereo ceramic cartridge phono I think the current draw was around or under 1mA.

That said could I also add a JFET buffer to the Mcintosh on its output and still get the maximum 10Vrms undistorted output it is capable of. I know in normal use it will never put out that much, but given it is capable of that the buffer needs to be able to do that. Now if I lose any gain from the JFET buffer I think that can be made up in the plate amp as I think it has an adjustible output for the high pass.

The concern about adding the buffer to the plate amp is it would have to be wired in between the input jacks I use and the board.

That would keep things simple and eliminate one box.

Another question.

The McIntosh is capable of a maximum undistorted output of 10Vrms and the plate amp has both RCA inputs and balanced XLR inputs. Which would it be best to use far as the overall gain structure is concerned and the maximum voltage handling without dfistortion?


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Tue 06, 2018 2:40 pm 
Member
User avatar

Joined: Jul Mon 26, 2010 8:30 pm
Posts: 23333
Location: Annapolis, MD
Quote:
That would keep things simple........

You're already way past that point.......;)

You can certainly make a buffer using a JFET in a source-follower configuration but, an op-amp IC might be the better choice. Could you translate the prose into a few specs?-- eg voltage swing, power output, output impedance, linearity, frequency response, etc.?

_________________
-Mark http://pixellany.com

"It's always something". --Gilda Radner (1946 - 1989)


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Tue 06, 2018 4:30 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19018
Location: Warner Robins, GA
For the preamp output the maximum undistorted voltage the preamp can put out is 10Vrms.

Impedance needs to be low maybe 10-20K

For the record out it puts out 1.2Vrms with 10mV at phono input and 200mV in its other functions with the rated input of 200mV into a 25K load. I maybe could get by with no buffer, but in phono and tape head mode the record output is right off the preamp stage which also has part of the compensation network connected at the same point so I'd be concerned that a lower load impedance would affect the compensation plus a .1uF cap driving a 25K impedance starts rolling off the bass at 100Hz.

I cannot find a spec for the output impedance of the main outputs, but if I take the .47uF output cap which feeds the 60K portion of the volume control I get roloff starting at 10Hz. Now if I take a 20K load and use that I start to see roloff at 40Hz.

That said even with no load connected I get a 10 degree phase shift at 30Hz.

I'm not exactly sure how a phase shift would affect the audio.

That said it would take a 2uF cap in place of the .47uF to get no phase shift at 20Hz. Thing is do I leave it alone and trust the engineers or do I fix the phase shift?


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Tue 06, 2018 6:37 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 533
Location: Equinunk PA 18417
Gosh I got confused by the first post. Are you thinking to strip down the crossover and repurpose it into unity gain buffers with no eq, for use outboard of the Mac, or are you thinking to steal some of the Mac's already-limited supply current for 4 more buffer stages? Do you have around 22+ volts available from the Mac for those buffers so they can pump out the max 10 volt peaks? It is so much easier to get what you want with a decent opamp. You don't need much gain, probably only unity gain, repurposed TLE2064 will do fine. Spec sheet says quiescent current for a quad package will be about 1.5ma, which would obviously go up with increasing audio output if connected to a super-low input impedance like a transformer.

A single quad package can do the job if you can tolerate a tiny bit of crosstalk between channels, mostly due to proximity of other circuit components. Personally, I'd be tempted to do that. Just make sure the power rails are well bypassed at the chip.

Personally,I don't find that sort of phase shift audible but I am not a "golder ears" type. 3db down at 10 hz is fine by me even if there is some phase shift starting well above that. Heck, there is phase shift through air. If you increase the output cap, don't forget to look for sub-audible or near-DC "junk" that may come with it.

_________________
-Richard Fairbanks


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Tue 06, 2018 6:56 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19018
Location: Warner Robins, GA
The initial idea was to do away with the crossover and repurpose one of the quad OP-AMP chips as a pair of buffers, but if it could be done in the C-24 (main B+ 75Vdc) I would prefer that as it would eliminate one box and me having to remove parts off a PC board.

Agreed about the cap increase.

The preamp itself has always sounded good as is.

Suppose I could disconnect it from the crossover then put my scope on the output and use an audio generator then see if I notice any phase shift.

I'm tempted not to second guess the McIntosh engineers though. Even if I solved the phase shift at the output I'd have to check every stage and possibly do the same.

Now for the record out given the maximum output voltage is low I could use a pair of JFETs in the preamp given they would draw very little current. At least when I made the ceramic phono buffer they drew extremely low current running on 9Vdc.

I could then pretty much leave the crossover as is then take the output signal from the first stage of the crossover (is a buffer) without having to remove any components other than disconnecting one end of the two .033uF caps.



EDIT:

Here's what I thought of for the record out buffer. Ignore the power supply.

Attachment:
Ceramic phono buffer 3.jpg
Ceramic phono buffer 3.jpg [ 71.31 KiB | Viewed 869 times ]


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Tue 06, 2018 9:21 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 533
Location: Equinunk PA 18417
Okay, so we've set aside the old crossover box for another project/day.

Me, I'd leave the preamp circuit the way it is because I wouldn't want to pass along sub-audible voltage drifts into a subwoofer. Huge waste of power when the speaker can probably just barely shove out anything below 25hz. Whatever phase shifts there may be below 100hz, I'd ignore them. The speakers no matter how good are a far weaker link. Oh, and, a crossover introduces phase shift too,

C24 has a 75 volt supply? Wow, that's about double what I would have thought for a solid state preamp. Well, surely a couple of fets won't tax curent demands much as long as you don't expect them to handle a very low impedence load and I like your idea to completely isolate the tape outs from whatever circuit they were previously hung off.

You asked about RCA vs. XLR, which is best? Impossible to say without further information, maybe just use whichever input is more convenient. Probably the RCA, it sounds like. Sometimes manufacturer's design enough gain for the RCA input and then build a psuedo-balanced pad dconcoction to knock down the higher signal level into the same preamp as the RCA jack. Sometimes they run the XLR through a separate stage designed for balanced signal, and pad that down. I've also seen the other case, where they take optimize for the higher balanced XLR signals, and build a separate gain stage as sort of an afterthought for the RCA input to boost it up to the following circuit requirements. So, no way to know which is "best" to use without knowing what they did.

_________________
-Richard Fairbanks


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Wed 07, 2018 12:32 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19018
Location: Warner Robins, GA
Agreed about the capacitors.

Surely they had a way to accurately measure phase shift back then and if phase shift was an issue it would have come with larger caps stock.


Here's the plate amp I'm thinking about getting.

https://www.parts-express.com/dayton-au ... --300-8010

Here's the manual

https://www.parts-express.com/pedocs/ma ... manual.pdf

It has a 5 year limited warranty so I don't want to install a buffer on the plate amp or do anything to it such that the warranty is voided.

That is the main B+ before the regulator.

The RCA and XLR inputs in high input mode have a maximum input of 2Vrms.

My guess is the balanced input is impedance balanced only and not a true balanced input which would be easier to integrate with the RCA jacks.

Possibly the same with the balanced output as well.

EDIT:

Just thought about it and I don't think I can do a JFET because it won't have unity gain which won't necessarily work unless I get the tape loop switch replaced so I can bypass the tape loop when not using it.

I do want a buffer inside the C-24 though for the record out so I don't have to worry about cable capacitance.


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Wed 07, 2018 3:21 am 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 533
Location: Equinunk PA 18417
Dayton, nice feature set at that price, I can see why you want one. Doubtful the XLR inputs give you any benefit. Sure would have been nice if Dayton spec'd the input impedance. 10k is pretty typical these days but who knows?

If you can steal 2ma from the Mac's supply you have enough to power a TLE2064, quad opamp. Unfortunately, you'd have to DC isolate its in and out (capacitors) because I think you've only got a unipolar supply in there, right? Still, once the supply concern is worked out the opamps would solve your problem.

_________________
-Richard Fairbanks


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Wed 07, 2018 3:48 am 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 533
Location: Equinunk PA 18417
I was just thinking, Dayton's 2v peak input level is slightly inconvenient. Your preamp's 10v output is 16db greater than that. A "standard" -10dbV signal is approximately 300mv. The 6db pad increases -10dbV input level to -4dbV, about 300mv to about 600mv, peak clipping moves from 2v to 4. Will that matter in your setup? I also would note that signal to noise is about 98db A weighted. Figure the residual hiss will be audible in your speakers (assuming you use Dayton's dsp crossovers). If your listening room is really quiet you may hear it. Hopefully it won't matter (probably will be low level) or you can adjust preamp vs. power amp levels to keep hiss down without messing up your ability to blast out favorite tunes on demand. I was thinking about this because I've been in a similar situation.

_________________
-Richard Fairbanks


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Wed 07, 2018 1:23 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19018
Location: Warner Robins, GA
Here's some measurements

With an input of 5.76Vpp I have the volume control set to put out a 15.35Vpp signal at 400Hz both channels driven with the switch set to L+R MONO no load.

Current slowly varies between 17.6-18.04mA.

With no input current is a steady 17.84mA.

I measured the current by removing the B+ fuse and inserting my meter there which is before the regulator.

I've got a unipolar supply.

The preamp is only spec'd at a maximum of 10Vrms output before distortion. In normal operation with the spec'd input signal I will never see that output voltage, although when I had Oldwirebender help with the crossover circuit I had it so that it could handle the full 10Vrms signal.

I can get some TL-082 OP-AMPs so I might try one of those for the record out and see how much if any it increases B+ current.

I will just use the crossover box and use a larger coupling cap to feed the plate amp input.

What I will do is measure the voltage going into the main amp so I know what the gain of the HP output of the plate amp needs to be set to.


EDIT:

Here's the buffer circuit I thought of using for the record output.

Attachment:
Buffer.jpg
Buffer.jpg [ 33.57 KiB | Viewed 811 times ]


The circuit will be added in such a way that it can be easily removed without evidence it was there.


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Wed 07, 2018 3:33 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 533
Location: Equinunk PA 18417
You have about 2ma of current available before exceeding transformer rating. TL082 requires 3ma no load, which is already into overload, then figure another ma or two for audio output work. You can cut the transformer overload 'way back by using a TLE2062. Its no load current is under .7ma, so staying within your limits most of the time even with load. Circuit you posted looks right to me, and you can add a 15-20k resistor from each output to ground.

edit: by avoiding the Dayton dsp, you'll have a quieter system but you'll loose the convenience of variable crossover points and its ease in tuning the acoustic crossover between sub and mains. Crappy choice to have.

_________________
-Richard Fairbanks


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Wed 07, 2018 4:50 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19018
Location: Warner Robins, GA
Will have to look for one of those.

Don't know what the S/N ratio is of the Lafayette LA-375 amp I am using as my main amp nor do I know what the S/N ratio is of the C-24 preamp so odds are a little noise from the plate amp won't matter much



Think my best option is to keep the current plate amp and get one of these mini DSP processors as it can do everything I am needing in a small box. That would also be better as I have complete control over everything in the whole audio frequency range including EQ meaning I can tweak the whole system to the room. I can even use room eq software and a calibrated mic and have the MINI DSP automatically adjust to the room.

https://www.parts-express.com/minidsp-2 ... r--230-320

I would need this plugin.

https://www.minidsp.com/products/plugin ... -in-detail

With that OP-AMP you listed I could possibly have both buffers self contained inside the C-24 then feed the MINI DSP which will then feed the amplifiers.


Last edited by Tube Radio on Mar Wed 07, 2018 6:20 pm, edited 2 times in total.

Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Wed 07, 2018 5:29 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 533
Location: Equinunk PA 18417
I forgot to mention, you should put a 100-200 ohm resistor on the output of each opamp. Doesn't matter if before the cap or after. It will ensure the opamp remains stable even with long, high capacitance cables.

_________________
-Richard Fairbanks


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Wed 07, 2018 6:28 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19018
Location: Warner Robins, GA
Hmmmmmmmmmmm........ the plate amp and MINI DSP options won't be a viable option because neither can handle the maximum 10 Vrms output of the preamp.

For an input of 199.6mVrms I get an output of 1.205Vrms at 400Hz.

For an input of 1.658Vrms I get an output of 10Vrms at 400Hz.

Looks like I may be better off using the crossover box I have and eventually looking for a pro audio crossover with DSP that has a good S/N ratio which can handle 10Vrms input.

I know that I will never see 10Vrms output, but the preamp is rated for that and I feel like any DSP or crossover should be able to handle that as well.

I'm tempted to just use a JFET for the record out jacks.


EDIT:

Those voltages I listed are wrong.

I tried just bridging the right and left inputs together with the switch set for stereo and got more output voltage. When I switched the switch to stereo I saw the output voltage jump up. I then just shorted the left input jack to the right input jack and the voltage stayed the same.

That said I also noticed I have a problem with the preamp in that one channel is lower in output than the other channel.

So before I do anything else I need to get that fixed.


EDIT 2:

Solved some of the imbalance issue.

I measured the 1K resistors I installed in the line inputs (preamp shorts to ground all unused inputs and some modern devices can be damaged by that) and they were all off a little (Radio Shack 5% carbon film) so I found some 1K metal film resistors and measured them then separated them in pairs as to which reads the closest to 1K. The differences between resistors were like I think several 100 milliohms. The resistors drop 3mV of audio signal across them, but that isn't enough to be noticeable.

The channel imbalance is different at different positions of the volume control so I measured each section of the volume control and they weren't exactly perfectly equal with each other at all volume control positions, but I don't think they were off enough to cause any real imbalance.

There's carbon comp resistors throughout the preamp so later on I may replace those at some point as a drifted resistor can certainly cause a channel imbalance issue.

I think that I have a couple JFETs so I may try them in the record out circuit provided I have some 1 meg resistors just to see how well they work and how much signal voltage is being dropped. If the signal voltage is being dropped by under 100mV then that will be ok. I'm not exactly sure, but I think that JFET preamp actually has some gain to it if I remember correctly.

EDIT 3:

Here's the data on that JFET buffer

Tube Radio wrote:
Output of rectifier 9.78Vdc .7mVac ripple
Second filter cap 8.86Vdc .2mVac ripple
voltage on 3.3K resistor 1.436Vdc
voltage on other 3.3K resistor 1.538Vdc
B+ current 920uA


So it looks like that circuit will definitely draw way under 2mA. All I need to do is find a source of B+, but I'm not sure what voltage I need.


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Wed 07, 2018 9:18 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 533
Location: Equinunk PA 18417
I don't see the advantage of single jfets except that you have parts on hand. Breadboard one of those puppies first and measure the current draw with full signal, as well as max voltage output. It looks to me like each one could draw a couple of ma at full signal into a load. Also, what's the output impedance? You want it low. But, that's only my chairside observations. Opamps aren't as much fun to play with, either.

_________________
-Richard Fairbanks


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Wed 07, 2018 9:31 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19018
Location: Warner Robins, GA
I think the lowest impedance I will see is 10K which is the input impedance of the USB sound card.

I will try the JFET just because I have them. If not satisfactory I will go with the OP-AMP.



EDIT:

Here's my solution for the record out buffer.

Attachment:
Buffer 2.jpg
Buffer 2.jpg [ 30.26 KiB | Viewed 752 times ]


B+ no load is 9.584Vdc as I tried to take the voltage from a regulated point that was lower in voltage to keep current low.
With the OP-AMP connected B+ drops to 9.556Vdc.

B+ current both channels driven with a 400Hz signal and a 20K resistor on each output is 725uA.

With the volume control full CW it takes an input of 1.136Vrms to produce 10Vrms out.
The output to the record jacks is 1.126Vrms.
The maximum output voltage of the OP-AMP right before distortion is 1.742Vrms.

I will install the circuit and see how well it works.


EDIT 2:

Here's a picture of the installed circuit.

Attachment:
Buffer.jpg
Buffer.jpg [ 116.59 KiB | Viewed 748 times ]


I used some double bubble epoxy to secure the circuitry.

The buffer works quite well and has not affected the audio quality at all.

It's a good thing I used the 47uF caps on the output of the buffer as the DBX 200 program route selector I am using to connect my 8 track, reel to reel and USB sound card normally sends the record signal to all three record outputs. I ran the numbers using a 3.3K load given I took the worst case and paralleled three 10K loads. The record out is flat to below 20Hz. There's three buttons labeled line and copy and I can press a button and the record signal does not appear at the input jacks of the selected device so I can use that to send to only one or two.

If the switches are set right I can for instance copy a reel to reel tape to my computer while listening to my phone through bluetooth connected to the C-24, my 8 track connected to the route selector or any other input on the C-24. Also I can have up to three processors and if I get a Dolby noise reduction box I can connect that as well.

That box will enable me to record from any of the three listed sources or anything connected to the other inputs of the C-24.


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Sun 18, 2018 5:10 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 533
Location: Equinunk PA 18417
TR,

If the opamp is running with 9.5 volts supply it should be able to give you 6.2 volts p-p at worst case into a 2k load (my estimation from the spec sheet). You measured only a quarter of that voltage before onset of distortion but into a lighter 3k load. Something is wrong there. Did you measure full negative to full positive peak-peak? I would expect you to find at least 4.4v rms (sinetone 6.2v p-p) before clipping.

I still recommend a small resistor, say 100 ohms, in series with the line between each 47uf cap and output jack. Spec sheet also recommends a resistor. There is no harm and it will ensure opamp stability no matter how much stray capacity is attached. Good practice also suggests to put a high value resistor, say 22k-47kohms, across the outputs to ground. You can solder them at the record output jacks if you wish. Their purpose is to keep the DC voltage at zero on the output jacks, which ensures the 47uf caps are always charged (better for long life) and give less chance of pops at the buffer output when switching destinations.

_________________
-Richard Fairbanks


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Sun 18, 2018 6:54 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19018
Location: Warner Robins, GA
I get a maximum of 4.826Vpp

The maximum output is greater than the max output of any source I can feed it.


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Sun 18, 2018 7:31 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 533
Location: Equinunk PA 18417
Excellent!

_________________
-Richard Fairbanks


Top
 Profile  
 
 Post subject: Re: Buffer amp
PostPosted: Mar Mon 19, 2018 1:18 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19018
Location: Warner Robins, GA
I figured that as long as the undistorted buffer output was the same as or a little greater than what it takes for full 10Vrms output that I would be good.

I will look into the resistors though.


Top
 Profile  
 
Post New Topic Post Reply  [ 21 posts ]  Go to page 1, 2  Next

All times are UTC [ DST ]


Who is online

Users browsing this forum: No registered users and 5 guests



Search for:
Jump to:  
























Privacy Policy :: Powered by phpBB