Selmer Amp: Too much gain!

Burns, Selmer, WEM, and more

Re: Selmer Amp: Too much gain!

Postby juniormarbles on Fri Jul 28, 2017 8:13 pm

Mr. Foxen wrote:Gain and distortion are different things. What you have is a distortion problem, it is not inherent to those amps, so something is wrong with it.

I agree: I can live with the high gain, if it's this model's design. I cannot live with the ugly distortion.
juniormarbles
Peasant
 
Posts: 13
Joined: Sun Oct 02, 2011 10:17 pm

Re: Selmer Amp: Too much gain!

Postby pdf64 on Sat Jul 29, 2017 9:30 am

To test an OT, get the amp on the test bench, hook it up to a sig gen and suitable dummy load (eg >100 W 16 ohm) and monitor the output across the load with a scope.
Apply a suitable sine wave (200 to 1kHz, 50 to 100mVrms) to the main chn 1 input and slowly increase the volume control from min until the max clean output is reached, eg when the sine wave starts to get its top and bottom extremes flattened off.
Make a note of the Vrms output across the dummy load (expect ~22Vrms) and then carry on increasing so as to overdrive the amp fully.
During the above, monitor the scope for any unusual behaviour, eg sudden drop out of the signal, which may indicate an intermittent fault, such as a bad OT winding.
As the cathode resistor is only 6W, don't spend too long (eg >10 seconds) with the output squarewaving, as it may overdissipate.
My band:-http://www.youtube.com/user/RedwingBand
pdf64
Knight
 
Posts: 1243
Joined: Thu Dec 14, 2006 9:55 am
Location: Staffordshire UK

Re: Selmer Amp: Too much gain!

Postby juniormarbles on Wed Aug 02, 2017 9:22 pm

An update:
I always assumed that there were 2 8Ω Celestion Greenbacks wired in series in the cab. Wrong. It's two 16Ω speakers, wired in parallel. The farting distortion is pretty much gone after installing 2 8Ω Greenbacks wired in series (the amp calls for a 15Ω speaker load).

But I did something else that now seems to hold the amp back:
With the stock 250Ω cathode resistor and two healthy EL34, I initially got a whopping 30W plate dissipation*. The tubes did not seem to mind, there was no red-plating, and the amp, though distorting in an unpleasant way at higher volumes, was loud. But the universally accepted safe PD of 70% of 25W max = 17.5W was nowhere near. So I doubled the cathode resistor to 500Ω, and, while I am now getting 17W and 19.8W respectively for each tube*, with plate voltages of 436/440, and current draw of .039 and .045, the amp sounds a bit anemic.

Any ideas why Selmer chose a cathode resistor that pushes PD* for EL34 into the stratosphere?

What to do? Will I stress the tubes by going back to the stock cathode resistor and the high PD (I am planning to install original GEC KT66 when I have a halfway healthy amp that won't risk blowing up these classics.)

* false measurement- see update below!
P.S.: I checked all coupling caps and eletrolytics (including cathode bypass cap). All work within specs.
Last edited by juniormarbles on Thu Aug 03, 2017 4:47 am, edited 1 time in total.
juniormarbles
Peasant
 
Posts: 13
Joined: Sun Oct 02, 2011 10:17 pm

Re: Selmer Amp: Too much gain!

Postby pdf64 on Wed Aug 02, 2017 9:47 pm

70% is the guideline for fixed bias.
Cathode bias requires hot AB conditions, which usually ends up meaning 100%.
25 watts plate + 5 watts screen grid = 30 watts
If, as you've found, you try to bias colder, it will tend to sound horrid at high signal levels due to the cathode voltage increasing, causing crossover distortion and constricting the power output.
See the section titled 'What about cathode-biased amplifiers?' at http://www.aikenamps.com/index.php/the-last-word-on-biasing

With the 250ohm cathode resistor, what are your HT and cathode voltages Vdc at idle, and mains and 6.3V heater Vac?

Are you sure that the HT caps C8 and 9 are good? if they're weak you may get nasty ghost notes at high signal levels.
My band:-http://www.youtube.com/user/RedwingBand
pdf64
Knight
 
Posts: 1243
Joined: Thu Dec 14, 2006 9:55 am
Location: Staffordshire UK

Re: Selmer Amp: Too much gain!

Postby juniormarbles on Wed Aug 02, 2017 10:16 pm

Here are the voltages, with original 250Ω cathode resistor:

Cathode resistor: 258Ω
"mains"/ secondary of input transformer, as measured at pins 4 and 6 of rectifier tube: 375VAC / each
B+ measured at output of rectifier: 454VDC
B+ measured after choke/center tap of output transformer: 449VDC
Cathode voltage: 32.1VDC
Center tap to plate, outer tube: 150.1Ω*
Center tap to plate, inner tube: 166.6*
Voltage drop across outer: 7.2VDC
Voltage drop across inner: 9.9VDC
Plate current, outer : 0.048A
Plate current, inner: 0.059A
Plate voltage, outer: 412.2VDC
Plate voltage, inner, 408.2VDC
Plate dissipation, outer: 19.8W
Plate dissipation, inner: 24.3W

This all looks pretty healthy!

*an authoritative "YouTube" gent said to measure resistance between plate pin 3 and rectifier pin 8 to calculate voltage drop and plate current. But he forgot there's a choke in-between. The added 50Ω between center tap and plate falsified current measurements and dissipation calculations.

Question No.1: is the difference in plate dissipation and plate currents between the two tubes significant enough to install individual cathode biasing for each tube and/or try to find better matching tubes?
Question No. 2: will equalizing plate dissipation improve sound or performance of the amp? If so, should I increase dissipation of the outer tube (19.8W), or reduce it on the inner tube (24.3W) ? Which way towards tube health and good sound?

P.S.: I tested C8, C9 with a Fluke capacitor tester, and both sections of this recent replacement cap read ca. 33mfd. Any further need to test for "weak"ness?
juniormarbles
Peasant
 
Posts: 13
Joined: Sun Oct 02, 2011 10:17 pm

Re: Selmer Amp: Too much gain!

Postby pdf64 on Thu Aug 03, 2017 12:43 pm

Yes that's all looking good, it would be reasonable to just leave it at that.
If you have a yen to tinker further, independent cathode resistors/cap for each power tube would be beneficial, as the self regulating nature of cathode bias can take full effect, leading to the tubes tending to 'match' themselves. I suggest 2 x 1k 3watt metal oxide and 47uF 100V cap in parallel for each cathode, leaving some air circulation room between cap and resistors, ie don't bundle them together.

To check C8 and 9 better, set the meter to Vac and see what it measures across them (ie between the + and - terminals of each cap), at idle and at high signal level. What model Fluke are you using?

Also the actual mains (240V) and heater (6.3V) Vac measurements are useful anytime you do any testing.

The power amp schematic has voltages noted on it; I suggest to go through your amp and measure each point, noting it down on a print of the schematic. Report back any significant discrepancies, eg >5%, ideally scan and upload your annotated schematic
http://www.vintagehofner.co.uk/selmer/s ... tosel.html
The V4b cathodyne voltages are particularly important.
Also check for any Vdc at control grid terminal 5 of the power tubes.

The suggested circuits for KT66 tend to have higher p-p OT primary impedances than those for EL34. It would be prudent to check what the OT in the Selmer is before proceeding. Hopefully it's up around 6k, but ~3k4 is commonly used with EL34 in fixed bias, so best to check.
To do that see http://www.regiscoyne.com/tech/unknownOT/
Make sure you understand things clearly before connecting one transformer to another! The OT needs to be completely removed from circuit beforehand, terminals isolated and insulated.
My band:-http://www.youtube.com/user/RedwingBand
pdf64
Knight
 
Posts: 1243
Joined: Thu Dec 14, 2006 9:55 am
Location: Staffordshire UK

Re: Selmer Amp: Too much gain!

Postby juniormarbles on Thu Aug 03, 2017 3:53 pm

Super helpful-thank you for holding my hand through this!
pdf64 wrote: I suggest 2 x 1k 3watt metal oxide and 47uF 100V cap in parallel for each cathode.

Would using the same 250Ω for each tube address the imbalance between the tubes' dissipation?
Should I not change and fine-tune one of these resistors to get matching current draws and dissipations? If so, which one (see my question No.2, previous post)?
Or does the change to independent cathode biasing achieve that by itself?

P.S.: All voltages are spot-on as per schematic. I am using a Fluke 287.
juniormarbles
Peasant
 
Posts: 13
Joined: Sun Oct 02, 2011 10:17 pm

Re: Selmer Amp: Too much gain!

Postby pdf64 on Thu Aug 03, 2017 5:31 pm

250 ohms is for 2 Power tubes sharing. With independent cathode bias for each tube, because the total current is ~halved, the value will need to be doubled to ~500 ohms, eg 1 x 1k in parallel.
I strongly suspect that when independently biased, the amazing powers of cathode biasing's self regulation will result in both tubes achieving closely similar cathode currents / plate dissipations at idle.
My band:-http://www.youtube.com/user/RedwingBand
pdf64
Knight
 
Posts: 1243
Joined: Thu Dec 14, 2006 9:55 am
Location: Staffordshire UK

Re: Selmer Amp: Too much gain!

Postby juniormarbles on Mon Aug 07, 2017 6:08 pm

Thanks again for your valuable information, especially that provided by PDF64 who held my hand through the whole process, including communicating offline, until all my problems were identified and solved.

Installing individual cathode biasing for the power tubes (2x 500Ω + 2x 47mfd) not only equalized the plate current (55 and 57mA*), but brought plate dissipation into the sweet spot: 23.4W and 24.0W respectively.

Through the individual biasing, the ugly distortion which was partially caused by mismatched speaker impedances, is now completely gone. I can now have a sweet, clean tone at low guitar volume, then turn up the guitar and the amp just sings in overdrive.

But what was most encouraging in this process is personal growth: I overcame my lack of confidence working on amps for which I used to rely on "amp specialists" who fixed things but did not empower me in the process. I am grateful for this forum that started me on the journey!

* P.S.: I find the calculated idle plate current awfully high, at least on paper. How did I end up with plate dissipation in the healthy range?
Here is my calculation (triple-checked, to make sure):

Resistance, CT to plate, inner tube: 160Ω, outer tube: 144Ω
Voltage drop across inner: 9.2V, outer 8.0V
Plate current (V/R) Inner: .0575A, outer: .0555A
Plate voltage, inner: 418V, outer: 422V
Dissipation (PC x PV): 24.0W inner, 23.4W outer

Did I miss something to get such high plate currents for EL34 rated for healthy operation around 35-37mA?
juniormarbles
Peasant
 
Posts: 13
Joined: Sun Oct 02, 2011 10:17 pm

Previous

Return to Miscellaneous British Amplifiers

Who is online

Users browsing this forum: No registered users and 3 guests