Pack-Bell DPA 30-1 6bq5

[johnblue]

Hello all, i have had this Packard bell DPA 30-1 for a while, i pulled it from a trashed console and all in all it was in working order. . i changed the coupling caps, and that's it... sounds pretty good, but a little "dirty". its time for an overhaul... i had another project that just got put on hold, so i might as well go balls out with this little guy. i've read its almost the same as the magnavox amps everyone loves, and i know that theres a few "famous" overhauls of this amp....

The DPA 30-1 is a little different from the more common DPA 30-3, in that it uses a PI filter with a choke... and the input jacks have coupling caps and a resistor, other than that its nearly identical to the DPA 30-3

its on the bench now and today i ordered some new rca and speaker jacks, some new filter caps.

There's something that i am going to do that most people advise against, and its disconnect everything from the 9 pin connector that powers the tuner/preamp.... so i am going to have to rework the power supply.

maybe you guys can help me with that... also there's a couple theoretical questions that i have regarding the topology as well.

i cant find a 250uf 25v cap for that area circled in the center that couples the two sides of the amplifier, im not exactly sure how that part works? is it a filter of some sort,? and can i up the capacitance a bit? can you guys explain how this area functions?

i put red x's where i took things out of the circuit, its not really running hot or anything, but im going to be removing the large can cap and using caps on a perfboard, so, im hoping to make some improvements, also, i have a few larger chokes im going to try.

 

[johnblue]

how do you post a decent picture?

 

[century tek]

My zoom doesn't want to work at the moment, but I do know that you do not want to remove the cathode bypass capacitor & resistor you have circled in red in the center.Without that, you won't get any sound.

 

[johnblue]

Thanks, I am not trying to get rid of that, I do have to replace it, I didn't know the proper name of that circuitry, so thank you now I can research.

The bypass capacitor is common in between the two sets of PP 6bq5,

The cathode resistor it 1200ohm 5w,
The stock capacitor is a huge old paper elcap,
250mfd 25v.

I'm wondering how to optimize this.

 

[gadget73]

If you want to post a big picture, you'll probably have to link it somewhere. The forum resizes things.

Leave the common cathode resistor. Its good for things overall. If you want to optimize it, consider converting it to using EFB to improve power output and distortion performance.


1200 ohm for a cathode resistor sounds unlikely. 120 is more what I'd expect. Cap values are actually reasonable. I might make it a 50v cap but otherwise I wouldn't screw with it much if the tubes end up at the right operating point.

Phase inverter appears to be a cathodyne (concertina, split load, whatever) type, which generally has good balance and performance. Also not particularly a concern. I can't read any of the component values, but from what I can see this appears to be a decently designed amp. The CLC power supply is extra nice, and not something you usually got in console amps. Some had chokes, but usually only the driver stage and maybe the screens ran through that. This looks to have the plate supply filtered by the choke, which is a very good thing.

 

[Retrovert]

The cathode-bias resistor moves (biases) the cathode into a linear operating range. Sharing the resistor keeps the total current through the resistor constant.

This resistor, however, introduces a form of Local Negative Feedback (LNFB) commonly called “Cathode Degeneration”. (“Degenerative Feedback” reduces output as opposed to “Regenerative Feedback” which increases output. )

Well, that's clearly bad. So what to do? Bypass the resistor with a capacitor! Because the capacitor readily passes AC, the fluctuations in voltage (the input signal is AC so the fluctuations are also AC) pass directly through the capacitor, bypassing the resistor's LNFB, and making up for the voltage drop. This eliminates the LNFB.

The problem is that a capacitor introduces harmonic distortion, phase shifts, and non-linearities because of reactance. That will fortunately be second order, which the ear finds pleasing. This is why bypassed cathode bias sounds "warmer" than unbypassed bias. (Tell that to your "tone controls are evil" friends who love the sound of bypassed cathode resistors!) This is why some prefer other bias techniques.

Because the bypass capacitor is in the signal path it affects the sound, and so it must be a high quality capacitor. The size of this capacitor also controls the LF response. Guitar amplifiers use a smaller capacitor because the guitar's lowest frequency is much higher than that for HiFi. The size of the capacitor is directly related to cost, and quality is also related to cost.

Do not tinker with any of these values until you understand how all of this works. You likely don't need to ever tinker with them. Dont' listen to people who "optimize" amplifiers without having a basic grasp of the principles and formulas (basic algebra) involved.

Enhanced Fixed Bias (search for EFB on AK as it was developed by DCGillespie here) , BTW, only works with ultralinear. EFB II is for pure pentode. This amplifier does not appear to have ultralinear taps. EFB and EFB II are a different bias technique. Search for "Fixed Bias" to understand how this works. EFB and EFB II do not have the harmonic distortion from cathode bias and bypassing.

I'd suggest you first replace every single electrolytic capacitor, every single carbon-composite resistor (they've drifted in value, had horrible properties when new, and have no mojo, no matter what anyone claims) with 1% metal film, and the ceramic capacitors (drift in value from age and are hygroscopic) with 1% film, and get it working using the original cathode bias. THEN you can start making mods like EFB II.

 

[johnblue]

Thanks for all the great info!
That's pretty much what I'm doing, but my local electronics store doesn't have a 250uf cap, so for now I put a 330uf 50v. Everything sounds okay.

If I am understanding correctly, this will have slightly less HF roll-off, but in theory be more susceptible to oscillation?

I ordered the correct cap as well. It's in the mail.

 

[biscuithead]

 

[Retrovert]

Thanks for all the great info!
That's pretty much what I'm doing, but my local electronics store doesn't have a 250uf cap, so for now I put a 330uf 50v. Everything sounds okay.

If I am understanding correctly, this will have slightly less HF roll-off, but in theory be more susceptible to oscillation?

I ordered the correct cap as well. It's in the mail.

Remember, the higher the value the greater the potentially for LF oscillation that can damage the amplifier.

The general rule is that the reactance of the cathode-bias capacitor at the lowest frequency to be passed should not exceed 10% of the value of the cathode-bias resistor.

So if Xc = 15 Ω, you're getting close to the limit which you'll hit at 470 µF (17 Ω).

You can calculate it for other capacitor values:

Xc = 1 / (2 × Π × f × C)​

Since Xc varies as 1/C, all things being equal, you can just compare 1/C1 to 1/C2. So the difference between 250 and 330 is a 25% decrease in Xc.

I'm not saying you'll have a problem, I'm just saying this is why bigger bypass capacitors are problematic.

 

[gadget73]

Honestly the stock 250 is on the large side. Most of the time those cathode caps were closer to 100uf. 220uf would also get the job done, since 250 isn't a standard size anymore.

 

[Retrovert]

Honestly the stock 250 is on the large side. Most of the time those cathode caps were closer to 100uf. 220uf would also get the job done, since 250 isn't a standard size anymore.

'zactly.

I looked up my calculations performed when I was working out the bias values.

Here's a portion of the numbers I had calcualted, so I can quickly copy to here:
25 µF @ 20 Hz is 319 Ω, @ 1 kHz is 6 Ω, and @ 10 kHz is 0.6 Ω
100 µF @ 20 Hz is 80 Ω, @ 1 kHz is 1.6 Ω, and @ 10 kHz is 0.16 Ω
470 µF, @ 20 Hz is 17 Ω, @ 1 kHz is 0.3 Ω, and @ 10 kHz is 0.03 Ω

So it is very clear that higher values start to present problems for LF stability. I'd say that 100 µF to 220 µF would be perfect and you'd never hear the difference between that and 250 µF. But, as I previously noted, I would go no larger than this.

 

[eiraved]

Thanks for all the great info!
That's pretty much what I'm doing, but my local electronics store doesn't have a 250uf cap, so for now I put a 330uf 50v. Everything sounds okay.

If I am understanding correctly, this will have slightly less HF roll-off, but in theory be more susceptible to oscillation?

I ordered the correct cap as well. It's in the mail.

Have you gotten your Packard bell DPA 30-1 up and running @johnblue ?