Davis 201A PA Conversion

[Pdavis68]

Okay, so the Hammond conversion is on hold. I'm pretty sure the OT is blown. My original thinking was I could pull the OT from my Davis PA, but I had never fired it up and I didn't want to deprive it of an opportunity to serve as a guitar amp.

I had never even taken the case off the chassis and was surprised to see that, despite the missing power tubes (which I had noticed), it had the 6CA4 and all 3 6EU7 preamp tubes. So I replaced the power cord (someone had yanked the third prong from the original) and threw in the 6BQ5s from the Hammond.

I've attached the service manual and a photo of the internals prior to adding the power cable.

Some notes regarding the internals: It's had some work done on it before I got it. The 3 old fashion mic jacks have been replaced with 1/4" jacks. Lots of shielded wire. The shielding on the input jacks has been cut off from the ground connector on the jacks (it appears to have previously been attached, as there are blobs of solder on each ground connector). The shielding IS grounded on the grid sides, though.

I'm guessing the ceramic caps were also after-market replacements. They all have bare wire connecting them and the bare wire is covered in heat shrink (notice most of the cap legs are covered in heat shrink).

So I fired it up going through a 40W bulb and it worked like a charm. Way early breakup on all channels (notice channel 1 is cathode biased, but 2 and 3 are grid leak biased), though. So I messed around a bit and then put in the 200W bulb and fired it up and it was a bit better, but still breaking up really early. About 4 on the volume.

You'll notice on the schematic, on the primary side there's a switch. This is the "tube saver" switch. When on, it reduces the voltage by either 10% or 15% (the manual says 10%, the sticker next to the switch says 15%). This applies to heaters as well. Anyway, I forgot that was on (I figured first time turning on that was better).

So I turned it off and then tried it again and it's much better. It starts to break up at around 6 on channel 2 (a little earlier on channel 1, I think).

It still doesn't have a great deal of clean headroom, though, so I'd like to address that first.

I still haven't put in grid stoppers on the inputs. The grid leaks are 2.2M which seems pretty high. Also, I don't think the tone stack is going to work out. I have to have the bass turned all the way down or there's just way too much mid and bass and it's really muddy. It's probably a good candidate for a Baxandall since it's got the 2 knobs. Never used one before and I've heard good things...

I'd appreciate some pointers on where to start first, in terms of looking for more clean headroom.

[olddawg]

IMHO.... There is not a lot of clean headroom any any 2x EL84 guitar amp, especially using a guitar with humbucking pickups.

[Pdavis68]

I was actually just about to update this and say that I discovered a lot more clean headroom when I plugged in my strat (I had been playing with my Les Paul). :-)

So looks like the appropriate clean headroom is there.

[Pdavis68]

I don't have a lot of experience with tube amps. I've played a few a bit, but the only one I've owned is a little Japanese guy with an SE 6GW8 (I now have the Davis, so my first EL84 based amp). So is it just hard to get a clean sound at volume with EL84s and humbuckers? I still haven't played it much and I have to go to a birthday party with my daughter for a bit, but I haven't tried controlling it with the guitar volume and really working the amp, though I will.

I'm curious, the amp has 4, 8 and 16 ohm taps as well as 25V and 70V taps and a "Booster Output" which comes out after the tone stack. Would any of those be appropriate for headphones (I'm not expecting anything particularly close to what it sounds like going through a guitar speaker), just something to noodle around on while the wife and daughter are sleeping.

[Justin Thomas]

I think the "Booster Output" is basically a "Preamp Out" in this case. Same as a guitar amp, run it into a larger power amp if needed. I doubt any of those outputs would be usable for headphones - not modern ones, anyway... and unless it's stereo, you'll only get one side in the phones.

Somewhat relevant question: what ARE those 70V output taps for? Is that for driving a field-coil speaker or something? Just wondering what they were originally intended to power...

Justin

[Pdavis68]

The 70V output is for long distance runs. Low voltage, high current loses power very quickly over long distance (I assume it's the same idea as high tension lines for running electric power long distances). Since it's a PA system, it's intended for the possibility of handling far away speakers. I don't know what kind of speakers are used for those, though. Perhaps it goes into a step-down transformer first.

[DRH1958]

You're probably not going to get any clean headroom form inputs 2 and 3 since they are grid leak bias. They historically have less than 1 volt bias, so they will be biased very hot. The top input is somewhat different than a normal cathode biased stage since the 6EU7 is basically the same tube as a 12AX7 and they only have a 560 ohm resistor for the cathode bias. Since they have a 2.2 meg grid leak and the 560 ohm cathode resistor, it almost appears they are using combination of grid leak and cathode bias. I usually see 1.5k to 2.2k for this position for normal cathode bias. If you do change this to strtictly cathode bias, then R1 will need to be changed to 1 meg because this is the highest permissible size for these tubes(12AX7 &,6EU7) Maybe measure all three input tube's grids to see what bias voltage you have. On ch 1 if your cathode is less than a volt, I would go to a value that gets you close to 1.5 volts and use this channel. The other two could be measured as well to see what the grid leak bias gets you, but it will probably be less than 1 volt. You could use these for your hot channels.

[Pdavis68]

Yeah, I think I'm going to have to change them to cathode bias.

It's got some sort of frequency-dependent static and I think maybe there's some oscillation involved. I haven't checked it with my strat, this was just with the Les Paul. My daughter is going to sleep, so I can't be playing it right now to test. I'll have to check tomorrow. But if I play low E and a B on the A string, together it's particularly noticeable. It doesn't really show up for for the lower D, G, B, and high E strings. Anything down low-ish on the E and A strings. I think the 1-3 of E & B resonates particularly and causes it to be worse. Before I added the grid stoppers, a couple of times, as the notes would start to fade out (it would be staticy as this was happening), it would suddenly cut to silence. Then if I hit a string after that, there'd be a momentary pause (I figure 50-100 milliseconds) and it would go from silence to whatever I played, like it had just been turned on.

It also seems to get worse after the amp has been on for 3 or 4 minutes. I guess I'll start thinking about my cathode bias options, as I'll want to use different values for each channel. I might do 1 LED bias channel, do one cathode biased with a bypass cap and one without the bypass cap. That should give me some differences.


I'm wondering if this is caused by the grid leak bias...

[DRH1958]

The 70V output is for long distance runs. Low voltage, high current loses power very quickly over long distance (I assume it's the same idea as high tension lines for running electric power long distances). Since it's a PA system, it's intended for the possibility of handling far away speakers. I don't know what kind of speakers are used for those, though. Perhaps it goes into a step-down transformer first.

You are correct, sir! The speakers actually have the output tranny mounted right on the speaker and there are taps on the tranny for volume control. This is/was popular in offices or schools when they needed to put multiple speakers covering all parts of the floor for announcements, muzak,etc. The speakers were mounted in the ceiling or walls.

[DRH1958]

I don't know if this will fix anything but there aren't any grid stoppers going to the inputs of your 6BQ5s. Try 10k or so. You shouldn't lose any high frequency response because the input capacitance is very low on pentodes, so this can help improve power tube distortion. Also on the output stage, there is no screen grid resistors.
I'm thinking that converting the input stage to normal cathode bias will help quite a bit.
And this came up in another thread. Your amp has the first three stages on the same B+ node. It was suggested that the first stage should be de-coupled from the next two stages and have it's own node for the B+ supply to it. It helps keep a copy of the signal being transfered to the other stages via the B+ rail causing oscillations. See post #57 of this thread, the middle paragraph.
http://music-electronics-forum.com/t39692-2/

[Pdavis68]

Good catch on the B+ being shared by the first 3 stage. Thanks! I'll add the grid stoppers to the power tubes as well.

For the screen grid resistors, what would you suggest? Would 100ohm 1W do the job? (That's the lowest value I have in the 1/2W & 1W arena).

[Enzo]

70v and 25v are what is called constant voltage speaker distribution. PA amps were used for PA systems, which usually means multiple speakers. If you have 20 ceiling speakers in a restaurant, imagine trying to wire them up for an 8 ohm or 4 ohm total. 70v output lets you just run a wire pair speaker to speaker across the ceiling, and tap across it at each spot with a matching transformer. the multi taps on that small transformer allowed you to balance the sound to suit. In other words, you might have most speakers set for 5w, but the one over the waitress station or the cashier might be set for 1w or a half watt so it is less loud. That ratio would remain regardless of the overall volume setting.

https://en.wikipedia.org/wiki/Consta...speaker_system
http://www.crownaudio.com/media/pdf/...nt_voltage.pdf

[Chuck H]

Something that hasn't been addressed...

I think you already know that none of the alternative outputs are suitable for headphones, but the fact that you asked compels me to state that you cannot operate the amp without a load. SS amps can be operated without a load or with a high impedance load, but do poorly with a lower than spec load. Tube amps are sort of opposite. They require the rated load and do ok with a low ish or high ish load, but no load may blow them up. A 600ohm pair of headphones would qualify as no load as far as a tube amp is concerned. So any notion you may have of safely running the amp silently with just a pair of headphones just won't work (for long).

HTH

[Pdavis68]

Thanks Chuck, I'm aware of that. I didn't think I was just going to be able to plug in a pair of headphones, but I thought maybe one of the lines might be close enough to 56 ohms (what my headphones are) that I could maybe throw a 10 ohm resistor over the 8 ohm output and set up some sort of circuit off maybe the 25V or 70V line to power the headphones.

[Chuck H]

Gotcha. You could always use that load resistor and make another resistive divider right off the + end with a high enough impedance so as to not affect the load the amp sees. IMHE though, this sounds like a pair of crows (caw caw ). I've even tried making a box with inductors and caps to simulate a guitar speakers EQ with only limited success.