Davis 201A PA Conversion

[Pdavis68]

Sorry if this has been done and I missed it, but have you measured the AC high voltage on the PT secondary? I'd think with the rectifier tube pulled you'd see if the PT was indeed 500vct or not. Then measure again with the full tube complement to get an idea of the actual load on the PT, and be able to predict the DC voltages more accurately.

Yeah, I'm going to do several different measurements when I get home. I'll do the transformer without the rectifier and get the unloaded transformer values. Then I'll get the rectified values without any preamp or power tubes and see what the unloaded voltages are. I'll also try with the 6CA4 from my Hammond amp and see if there's any difference. I'll also double-check the resistor values in the power supply and see if they've been changed at all. Though the resistors wouldn't account for the 50+ voltage difference from the first filter cap.

[DRH1958]

Isn't a transformer usually measured with the circuit operating under load? It seems the off-load voltage means nothing because transformers have different regulation percentages. Some 250-0-250 traffos might read 350v off load and some might be 30-40 volts either way of that number. Don't mean to be argumentative, just practical. And the specs given for a traffo will be the loaded voltage not the off load voltage. Obvoiously this one has no given specs, just illustrating the importance of voltages under load.

[Pdavis68]

My interest in measuring loaded vs. unloaded is in trying to determine why what I'm seeing in voltages doesn't match what the schematic shows. It may well turn out that the transformer has been changed, but who knows, I may pull all the tubes except the rectifier and suddenly see that 330V and 285V on the third and fourth filter caps.

[eschertron]

My interest in measuring loaded vs. unloaded is in trying to determine why what I'm seeing in voltages doesn't match what the schematic shows. It may well turn out that the transformer has been changed, but who knows, I may pull all the tubes except the rectifier and suddenly see that 330V and 285V on the third and fourth filter caps.

If you get an accurate measure of the filter resistors, you can calculate the current passing through each node, and figure what each node is supplying to the tube complement. Regardless of the ratings on the transformers, this can give you a picture of the health of the components, and whether or not they are operating within their envelope.

[olddawg]

If you want a dedicated headphone amp check out a little MS-2 Marshall. They are around $30, less on sale. You will be surprised. It has a speaker in it too. You can also go DI to a PA with it.

[Pdavis68]

If you want a dedicated headphone amp check out a little MS-2 Marshall. They are around $30, less on sale. You will be surprised. It has a speaker in it too. You can also go DI to a PA with it.

I actually have one of those AMPlug Vox AC30s that I use in my car (not while driving, of course. I play in my back seat during my lunch break at work). It's not too bad for what it is...

[Pdavis68]

If you get an accurate measure of the filter resistors, you can calculate the current passing through each node, and figure what each node is supplying to the tube complement. Regardless of the ratings on the transformers, this can give you a picture of the health of the components, and whether or not they are operating within their envelope.

Good point. I'll do that as well.

[Pdavis68]

Just an update. I'm updating the schematic with the actual values from the amp (and putting them next to the components where they should have been in the first place). I'm about 75% done with it and will hopefully finish it tonight and get it posted here. A lot of values have changed.

[Pdavis68]

So I finished updating the values on the schematic and I've uploaded it here. I added the note in the upper right for the benefit of people who might search for a schematic for a Davis 201-A and wonder why their component values are so different.

Mic 1 and Mic 3 swapped places.

My current plans are:

1> Add screen grid resistors (200 ohm, I think)
2> Change the preamp grid stoppers to 1M
3> Add another power tap below the 260V (maybe down to 250V with 10u cap) for the first stage preamps, so they're not sharing the power with the second and third stages.
4> Cathode bias the first stage preamps (well, I may LED bias one of them) and change the plate resistors to 100K. I'll bias each one different.

I'm going to remove 2 jacks and add 3 on/off switches, one for each channel. So in the on position, the channel is connected to the input jack. So you can basically parallel any of the 3 preamp channels (kind of like jumpering the second channel in a Marshall). But in this case you don't need the jumper cable and you can do any combination of the 3 channels.

I'll do things in stages. Make a change, test it out and see how it sounds, make the next change, try it out. Etc...

After all that is done, I'll start looking into changing the tone stack to a Baxandall.

[eschertron]

Looking at (R29) NFB resistor, I see it's 100k with 220R on the tail of the gain stage (R23). Might as well not have NFB with that ratio. What's the factory value for those resistors?

[Pdavis68]

Looking at (R29) NFB resistor, I see it's 100k with 220R on the tail of the gain stage (R23). Might as well not have NFB with that ratio. What's the factory value for those resistors?

Actually, the NFB was 100K originally, except that it was at the 16 ohm tap. They moved it to the 8 ohm tap. R23 was originally 1.5K. I might want to double-check that I got that one right because I was second-guessing that 220R value when I was updating the schematic. Maybe it was 2.2K and I misread it or maybe I was looking at the wrong pin. I'll check that tonight. I've been thinking of using a pot for the NFB. That might make it easier to clean it up when using my Les Paul.

[eschertron]

Maybe it was 2.2K and I misread it or maybe I was looking at the wrong pin. I'll check that tonight.

2.2k I'd believe
That and the 16R tap will get about 14x more NFB.

I modded a SF champ and added a pot on the NFB. IIRC the original was 2.2k NFB with a 47R tail. I put a 25k audio pot in series with a 220R resistor. dialed to one extreme, I get about 12dB NFB to make the amp quieter and a lot more polite, dialed the other way it's essentially no NFB. I use it primarily as volume, to match the acoustic environment. A rage monster it's not, but it goes from 'subtle and clean' to not-so-clean very well.

[Pdavis68]

I think 220R is the correct value for R23 when R22 is 470k as this will bias the directly coupled cathodyne PI correctly. If R23 is 1k5 or 2k2 the plate voltage will be too high to correctly bias the cathodyne. If you want to adjust (increase) the feedback keep R23 at 220R and reduce the value of R29.

The original value was 1.5K and the original voltage was 285V, instead of 260V and the PI was the same as it is now. The original plate resistor (R22) was 560K instead of 470K.

Also, I'd like to LED bias one of the channels, but I've read that LEDs aren't terribly linear below 5ma and a 12AX7 isn't going to run that much current. Is there a way I could steal some current from somewhere just to get the LED above 5ma?

[eschertron]

Yep. I agree. Goes to show how little I notice. Even with these values the follower is biased quite warm (about -0.7vdc by my charts). Any changes to the NFB should be to the NFB resistor R29 itself.

[Dave H]

You were too quick for me. That post of mine was rubbish (the part about using 220R for R23) so I've deleted it. I was thinking of the voltage needed to bias the cathodyne and forgetting about the saturation voltage of the DC coupled stage driving it.