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


It is currently Sep Wed 23, 2020 7:39 pm


All times are UTC [ DST ]





Post New Topic Post Reply  [ 10 posts ] 
Author Message
 Post subject: Amplifier build
PostPosted: Jan Sat 18, 2020 8:21 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20790
Location: Warner Robins, GA
I have a Magnavox AMP-142 amplifier that has been gutted except for the transformers, choke and tube sockets.

I've been wanting to build a mono amp with it, but haven't had the money for parts yet.

While dealing with a phase splitter idea, bb.odin turned me onto the following self contained phase splitter chip.

Attachment:
DVR134_DVR135_BlockDiagram.jpg
DVR134_DVR135_BlockDiagram.jpg [ 79.42 KiB | Viewed 1172 times ]


http://www.ti.com/lit/ds/symlink/drv135.pdf

It can do 17Vrms into a 600 ohm load which will give me 48.0828Vpp.

That can drive the quad of 6V6 tubes to full output.

So I'm thinking of using that as the phase splitter in a separate box with a volume control and a +/- 18Vdc supply plus whatever preamp stage I need before the phase splitter as it by itself has a higher input signal requirement for full output.

I will also use a single 5U4 instead of the dual 5Y3 tubes.

The power supply for the preamp will most likely be a Traco +/- 18 Vdc supply which will run off the 6.3Vac heater wiring once I rectify and filter it.

I would add a single pole multi position rotary switch and have a few stereo inputs bridged to mono with resistors.

I may also integrate a low pass filter section before the phase splitter with its own level control so that it can be used with my Philco 118 radio so that it has some bass response as I had to reduce some of the lower audio frequencies going to the original speaker since one of the spider arms is broken and I didn't want the cone to move much at all to preserve the rest of the spider.

The speaker will be a 15" Jensen and three Jensen tweeters I bought from a forum member who removed them from an Emerson console.

I'll need to have an open back cabinet built for the speakers, amp and preamp.

Thinking I may find a rack mount box with no holes in it and use that to contain the preamp, controls, switches and jacks.

Thinking I'll need at least three inputs. One for a bluetooth receiver one for a computer and one for whatever I desire, but I may add more depending on what switch I find.

A ceramic cartridge phono input may also be added with both left and right channels just shorted together and the proper load resistor used.

Now to eliminate any issues with treble loss at the higher audio frequencies I may use an OP-AMP buffer with the necessary gain (will make it adjustable) to drive the phase splitter to full output then use a lower resistance volume control such as a 25K control.


Top
 Profile  
 
 Post subject: Re: Amplifier build
PostPosted: Jan Sat 18, 2020 9:38 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 643
Location: Equinunk PA 18417
Quote:
It can do 17Vrms into a 600 ohm load which will give me 48.0828Vpp.
Not sure about this. It can swing each output to approx. 1.5v each supply rail, which means you'll get around plus/minus 16.5vp-p swing from each individual output, relative to ground, using a bipolar +-18v supply. Is that enough for your goal?

I used these in a balanced audio system many years ago, back when this ic was under the Burr Brown flag. It's a very good chip and easy to work with. Should work well as a phase splitter although I have no experience with in that way. Bypass both supply rails right at the chip like the spec sheet suggests and it'll give you no trouble.

_________________
-Richard


Top
 Profile  
 
 Post subject: Re: Amplifier build
PostPosted: Jan Sat 18, 2020 11:44 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20790
Location: Warner Robins, GA
I think how it does it is because each OP AMP can do to +/- 1.5 volts of the supply so the non-invert output is +/- 1.5 volts of the supply and the - is +/- 1.5 volts of the supply.

So that would mean the voltage between the invert and non-invert is double what a single output can do by itself.

Here's a calculation I did based on the datasheet maximum of 27dBu.

Attachment:
Calculation.png
Calculation.png [ 8.56 KiB | Viewed 1145 times ]


I can easily get an Edcor XSM series transformer to step up the signal should that be necessary. Those are rated to 50Vrms.

I recently used a phase splitter to step up the output of an FM tuner which only had a 6.3Vrms heater winding for power to derive a +/- supply only I used a dual OP-AMP to do it which was not optimum, but works. That was then fed to a transformer to make the output unbalanced. I was able to get 32Vpp before distortion on a supply of only +/- 8.5Vdc.


Top
 Profile  
 
 Post subject: Re: Amplifier build
PostPosted: Jan Sun 19, 2020 3:41 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 643
Location: Equinunk PA 18417
Quote:
It can do 17Vrms into a 600 ohm load which will give me 48.0828Vpp.

This is confusing but I see now your thinking. 17vrms for a sinetone would give you nearly 24vp-p, so if you had two 17vrms signals of inverted polarity, you could measure 48 volt peaks differentially between the two. Trouble is you won't get 17vrms from either ic output pin. Neither output can exceed the supply rail. You'll get, on a good day, 17vp-p from each pin. Count on 12vrms from each output pin.

When you said phase splitter I was thinking that the chips differential output is being used as two single-ended outputs, with inverted signal polarities of course. A 1:1 transformer does not change things at all because its secondary voltage will end up being split between the tubes and their common ground reference. A step up transformer does give additional voltage gain.

_________________
-Richard


Top
 Profile  
 
 Post subject: Re: Amplifier build
PostPosted: Jan Sun 19, 2020 3:42 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 643
Location: Equinunk PA 18417
double post deleted

_________________
-Richard


Top
 Profile  
 
 Post subject: Re: Amplifier build
PostPosted: Jan Sun 19, 2020 5:25 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20790
Location: Warner Robins, GA
The datasheet does show the 17Vrms maximum.

Perhaps the datasheet is in error?

Given a power supply of +18 and -18 that would mean the output could swing around 16.5Vp positive and 16.5Vp negative on each output pin, right?


Top
 Profile  
 
 Post subject: Re: Amplifier build
PostPosted: Jan Mon 20, 2020 2:52 am 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 643
Location: Equinunk PA 18417
Youre right the datasheet says 17vrms out! That must be differentially between the two outputs since it would be impossible for a single output with +-18v rails. So, differentially, that means each output is able to swing at least 8.5vrms,

Perhaps I was wrong to say each output can give you 12vrms, because that 17vrms figure would indicate considerably less. I wonder about it because the sheet guarantees the outputs will swing up to 3 volts away from the rails. (Actually each output can swing as low as 2 volts above the negative rail and as high as 3 volts lower than the positive rail, but since the datasheet is talking about an undistorted sinetone of 17vrms, we should assume that the output voltage swings within 3 volts of each rail.) The supply is +-18. So, I would assume each output can swing +-15v peak, which is 10.6vrms (not 12 as I posted earlier). Two outputs differentially should make 21.2vrms, so I'm not sure why the sheet states only 17vrms (into 600 ohm load). Maybe that is what you can expect with a more common +-15v supply but why didn't they just say that, and since when would the sales team allow such a conservative statement?

Anyway, if I had easy access to +-18 volts I'd be happy to wire up a DRV134 to check it out. I only have +-12v handy on my bench, not sure it is meaningful to extrapolate from that but I will wire one up and see what gives. Give a day or so.

_________________
-Richard


Top
 Profile  
 
 Post subject: Re: Amplifier build
PostPosted: Jan Mon 20, 2020 5:23 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20790
Location: Warner Robins, GA
What is going on I believe is this.

It's 17 Vrms into a 600 ohm load, but each output has a 50 ohm series resistor.

Now that would mean some voltage is dropped across the 50 ohm resistors.

For a 17Vrms output into 600 ohms that is a current of 28.33mA.

The voltage drop across each 50 ohm resistor would be 1.4165Vrms.

That would give a total drop of 2.833Vrms.

Adding the 17Vrms to that gives 19.833Vrms.

So unloaded, the chip should be able to do 19.833Vrms.

Now if I take a pair of 220K resistors and put them in series across the output to simulate what one might use for the output tube control grid resistors I get a current of 40uA.

Now the drop across each 50 ohm resistor is only 2 mVrms.

So i would have a maximum output of 19.829Vrms which is 56.0843436Vpp

That would definitely drive the quad of 6V6 to full output and then some.

I do plan on buying one Thursday along with the necessary parts to make a +/- 18Vdc power supply from a single 6.3Vrms winding. Now the benefit of the small Traco +/- power supply modules is that they are isolated which means I could use a pair of diodes to develop a + and - supply then feed that to the power supply module's input unless Traco makes one that runs on 8Vdc.

What I would like though is some sort of OP-AMP before that chip to provide for a higher impedance input without any upper treble loss as the 10K input impedance is fine for modern devices, but a lot of the older stuff would have a problem with that.

I figure on a dual OP-AMP with the first OP-AMP stage being a unity gain non-inverting buffer which will drive a lower resistance volume control of perhaps around 25K. That will then feed a second OP-AMP stage which will have adjustable gain so that I can set it to where the amp can be driven to full output with a typical line level source at nearly full CW rotation of the volume control.

Now with the OP-AMP receiving +/- 18Vdc supply and if I choose one capable of outputting close to the rails lets say within 500mV of each rail, that would mean I could input a max signal of 12.374Vrms before distortion sets in so the buffer will be fine without any volume control before it as I know of no unbalanced audio device I would ever use which puts out that much signal.

All I would then need is a box with a switch and multiple jacks.

The benefit there is absolutely no need of any coupling caps except maybe one on the input as some more modern sources have DC on the output.


Top
 Profile  
 
 Post subject: Re: Amplifier build
PostPosted: Jan Tue 21, 2020 1:07 am 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 643
Location: Equinunk PA 18417
I did a quick breadboard with DRV134 and +-12v supply. Very much like the sheet says, each output swings to about 2.5 volts away from the supply rails, in this case about 9.7V peak or 6.9Vrms on each pin before clipping. Differential connection doubles those numbers of course. When the differential output is loaded with 600 ohms, the outputs drop a couple of volts at which time each output pin shows very nearly unity gain. So, I think you can count on 17Vrms with a 600 ohm loaded output and higher if unloaded, assuming a full +-18v supply.

Both of the DRV134's I have are NOS from Burr Brown but I expect the TI releases should be identical. I noticed that the voltages at each output pin do not exactly match with a difference of about 2%. This was true with both chips I have here. Since they are supposed to be laser trimmed I did not expect a difference that large. For instance, when 5.0Vrms is applied to the input, one pin measures about 0.1v more than the other whether loaded or unloaded. Not a big deal and may not exist in recent product.

_________________
-Richard


Top
 Profile  
 
 Post subject: Re: Amplifier build
PostPosted: Jan Tue 21, 2020 1:30 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20790
Location: Warner Robins, GA
Thanks.


Top
 Profile  
 
Post New Topic Post Reply  [ 10 posts ] 

All times are UTC [ DST ]


Who is online

Users browsing this forum: regenfreak and 5 guests



Search for:
Jump to:  


































-->


Privacy Policy :: Powered by phpBB