Calling 6DN7 Experts.

[ToolGuyFred]

I listen to a pair of original, 2-tube monoblock ParaSEX power amps.

Very recently I have been plagued with pops and bangs of random loudness at random intervals from one amp. From the construction manual I did the tour of resistances and voltages at the test points without finding anything terrifying.

The physical condition of the output stage cathode resistor however, warrants more attention. This is a 2W tantalum film and has got hot enough to crack its outer shell which now exposes a hard but crumbly core. This test point also has a lower voltage than it should (more current, more heat, life expectancy is short).  I am going to replace this resistor and its partner cap with one for each tube (something I should have done long ago).

Fair enough. Roughly half the capacitance and twice the resistance for each tube will get me roughly the same operating point. Due to a purchasing mistake however, the output choke and output are the Magnequest set for the SINGLE tube ParaSEX. My understanding is that this gives me an output impedance higher than optimal for tubes operating at this current.

So I have the opportunity to re-bias my amps to operate more optimally with the hardware I am currently using.

Design value for existing cathode resistor: 390 ohm
Design cathode bias voltage: 22.5V
Actual cathode bias voltage: 18.4V

Power and bass response are not of huge importance as I listen with Lowthers + active sub in a small room at modest volumes.

Anyone for load lines?

Thanks in advance.

 

[Paul Joppa]

I believe we've talked about your amps before, but I don't remember many details so I'll treat this as a completely new thread. (It should probably be moved to Legacy Products, too - I don't know how to do that, but hopefully one of the other guys will do it.)

Actually, the ParaSEX was the fourth - and last - version of the 2-tube monoblock S.E.X. amps. And there were two variants of that. They are easily distinguished by the power transformer, the deYoung with many colorful leads, or the Magnequest PGP8.1 with solder tabs.

In both cases, the original plate choke has a secondary winding in the output cathode lead, providing a few dB of feedback. If you have the MQ choke, you have probably connected the cathodes directly to the bias resistor.

There are several possibilities for the pops and bangs, but I need to know what power transformer and plate choke connection you have before I can say very much more. The easiest one to check is to swap the tubes between amps, and see if the noises go with the tubes or stay in the amp.

By the way, one of the known issues is that cathode resistor, which should have at least a 5-watt rating, so I am actually surprised it has lasted this long. It is probably the wrong value anyhow, but I need to know which power transformer you have before I can comment further.

 

[ToolGuyFred]

Hi Paul

Thanks for your swift response.

It is the De Young transformer. On power up I get 500VDC on the HT, this drops to 450V when the heaters warm up.

Yes the Magnequest choke is just connected to the HT so I have lost the few db of feedback from the original circuit: yes, the cathode resistor is connected between cathodes & ground. Interestingly, while examining the cathode resistor, a piece flaked away  when poked with a multimeter probe and it's been silent since.

I have tried replacing the tubes with known good used ones without any significant change in behaviour.

The plan is to replace with a cathode resistor for each cathode, replacing the one 2W tantalum with a pair of 5W Mills. In the absence of any suggestions, I would replace the one 390 ohm resistor with two 820 ohm resistors.

 

[Paul Joppa]

The voltage is a little disturbing. High voltage at the power section plate (tube socket pin 2) should be around 370v for the deYoung, with the PGP8.1 it should be around 441v. Is your power line voltage excessively high?

The Magnequest plate choke for the stereo SEX amp is rated 40mA, so you'll want to reduce the current to 20mA for each tube. Bias is about -28v, so 1400 ohms for each cathode would be about right. Two watts is adequate for these, though I would probably use 3w-5w resistors myself. The cathode bypass caps should be rated at least 35v - the original was only 16v, which if it has not been replaced it is a likely source of noises (and will eventually explode, barfing its slimy guts onto the surface below). This operating point should be well-suited to the 8K impedance of the MQ output transformer you have on hand; I would nominally use 1.5uF parafeed cap, but the stock 3.3uF is still workable.

Here are some other things worth considering as long as you are under the hood anyhow:

* put grid stoppers on the power stage grids (tube pin 1). Anything from 100 ohms to 1000 ohms should be OK. Carbon composition is the old-school standard, but metal or carbon film seems to work well enough.

* replace the power supply caps - they are getting old and they are operating close to their voltage limit. That's another candidate for causing funny noises as they fail. Stock value was 10uF/450v; you can increase to 47uF. 450v is the minimum rating, but higher voltages are nearly non-existent unless you use film caps, which are comparatively huge and a pain to fit. Because of the marginal voltage rating, look for 105 degree C temperature rating and a long spec lifetime - say, 5000 hours.

Hope that helps!

 

[mcandmar]

Vishay make a range of 500v 105c rated caps that would be ideal.

 

[ToolGuyFred]

Many Thanks guys, very helpful.

My power line voltage is around 250V, it doesn't generally get higher in the UK.

I usually write the year on amps I build and these are 1999. In the spirit of experimentation, many different parts have been tried in these amps.

At present, power caps are a 650V 15uF polypropylene on the rectifier side of the choke and a 600V 47uF oily Mundorf bridged with a 1uF poly at the audio end. None of these are showing physical signs of stress. Cathode bypass is a Black Gate, sadly unreplaceable.

My shopping list then: Mills 5w for the cathode resistors, 0.5w 220 ohm Allen-Bradley carbon for the grid stoppers. I'll probably go with Elna Cerafines for the bypass caps.

Will report back when they're fitted.

 

[Paul Birkeland]

Yes, replace those cathode resistors ASAP.  When they fail, they open up, and then your cathode bypass cap will blow out after.

To reduce that B+ voltage, you can drop the value of that 15uF cap to get things better aligned. It would also be worth measuring the AC voltage that heats the 6DN7's, since your line voltage is a little high.

 

[ToolGuyFred]

Hi Paul

Thanks for your input.

Cathode resistors have arrived and will be fitted this weekend.

The 250V line voltage is within UK specifications but often equipment purchased from overseas "for UK use" is actually only rated at 240 or even 230V  (230V being common for mainland Europe). The step-down adaptor supplied with my Bottlehead DAC for example just can't hack more than 10 minutes without overheating & shutting down. (Not complaining though, the DAC sounds *much* nicer on NiCADs.) I'm not panicking too much over the HT voltage as these amps have been in continual daily use for the last 17 years (and 450V isn't a long way from the 441 I would have if I were using the PGP 8.1). 6DN7s last for years and I have only had two failures: The voltage divider for the heater bias failed on one amp and the other was a cathode resistor & cap traced to a NOS tube which "out of the box" was intermittently short between anode & cathode (ouch!). Will look around my bits boxes for a smaller pair of caps though.

I will double-check the heater voltages; I didn't write them down last weekend but I'm confident these are OK.

 

[ToolGuyFred]

Job done.

The heater voltages measured 6.31 VAC so nothing frightening there.
Amp #1 has HT running at 451 VDC and amp #2 at 453 VDC.
The cathode bias resistors are 2.7K and 3K in parallel which calculates to 1421 ohm (and measures 1380).
I didn't get anything like the 28V bias expected though. Amp #1 had 22.5 and 19.5. Replaced with new tubes and got both reading 22V. Amp #2 reading 20.5 and 21.5.

Pictures show before and after on amp #1. Comparison of cathode resistors - odd how one has failed so convincingly when they have both had the same number of hours at pretty much the same voltage. I expect the wire-wound replacements to be more robust, especially as I'm replacing one 2-watt resistor with four 5-watters.

Only just finished the work. Slightly less gain is making the sub sound louder (will tweak). Maybe a touch more upper treble detail. I expect this to change over the next few hundred hours, simply because things usually do.

Many thanks for your help, guys.