mike stehr Posted January 3, 2022 Author Share Posted January 3, 2022 On 11/9/2021 at 1:07 AM, captainbeefheart said: I had one of these a long time ago and ended up doing major surgery as well. I know you mention a long time ago, but do you recall having to re-bias the amplifier's output tubes at a lower dissipation rating? Quote Link to comment Share on other sites More sharing options...
captainbeefheart Posted January 3, 2022 Share Posted January 3, 2022 1 hour ago, mike stehr said: I know you mention a long time ago, but do you recall having to re-bias the amplifier's output tubes at a lower dissipation rating? Yes those poor tubes were screaming away at 15-16 watts plate dissipation. The plate voltage being too high was the main problem, higher than the schematic says which should be 345-ish. Even still the cathode resistor they chose is a little on the low side, I find 150 tends to work better for most EL84's which is slightly higher than 130 ohms most datasheets call for. So between lowering the B+ and using 150R on the cathodes they biased up much better. You're still near max plate dissipation but that's what this amp wants. I didn't increase that big 10 watt 100 ohm resistor like some people I read do because it sagged the amp down too much during testing so I went with a zener string on the high voltage secondary center tap. I was on the fence of adding bucking transformers but never did it. I know you gutted it and re-wired. I have found on many of these amps you sometimes get noise issues that can drive you nuts, it ends up being a poor ground connection from the tag boards because they use a rivet to mount them to the chassis and that grounded lug is what's used to ground the circuit. I drilled the rivets out, cleaned the oxidation and used hardware with lock washers and little loctite on threads. Quote Link to comment Share on other sites More sharing options...
mike stehr Posted January 3, 2022 Author Share Posted January 3, 2022 14 hours ago, captainbeefheart said: Yes, those poor tubes were screaming away at 15-16 watts plate dissipation. The plate voltage being too high was the main problem, higher than the schematic says which should be 345-ish. Okay, then that confirms my math for determining plate current from cathode voltage. You must have used the Sam's schematic with the plate voltage value. The original Pilot SA-232 schematic shows 320 volts on the plates. I have no idea why there is voltage differences between the schematics. The Pilot schematic mentions; "measure at 117VAC". But I'm sort of curious if maybe they measured at 115VAC? The Sam's schematic falls more in line with voltages stated measured at 117VAC from what I come up with. If a guy looks at the values for both Pilot and Sam's schematics, and determine total plate current from cathode voltage, they both just hammer on the 6BQ5s, if my math is right... You mention the amplifier (sounds?), or (operates?) best with the tubes at near max dissipation? I can just imagine back in 1960... "This amplifier sounds wonderful for a few months, and then the tubes glow red." Seems kinda rich even back when readily available 6BQ5/EL84 were 3-4 bucks a pop. Using a combination variac/isolation transformer, with 117VAC in, I get 360 volts B+ on the first cap, with 335 or so on the plates, and 12 volts on the cathodes. My numbers for dissipation of the output tubes are the same as you mentioned. I swapped in 150 cathode resistors, it raised the plate voltage 3-4 volts, and lowered the cathode voltage about a volt. It still comes out around 13-14 watts per tube. Keep in mind I'm trying to average things out, because I'm not using matched output tubes. I don't really want to cook my good 6BQ5s, so I'm using a junk quad to get going. They match emission-wise on a tester, but that's pointless in circuit. If I sub in a 5R4 rectifier, I get 312 on the first cap, around 285-290 on the plates, and around 11 volts on the cathodes. The drops the dissipation number down to 10 watts for each tube. Roughly 85% of the total 12-watt dissipation rating I'm guessing. It doesn't drop much voltage on the driver and phase inverter tubes, so it may actually pass a signal. Quote Link to comment Share on other sites More sharing options...
captainbeefheart Posted January 4, 2022 Share Posted January 4, 2022 1 hour ago, mike stehr said: Okay, then that confirms my math for determining plate current from cathode voltage. You must have used the Sam's schematic with the plate voltage value. The original Pilot SA-232 schematic shows 320 volts on the plates. I have no idea why there is voltage differences between the schematics. The Pilot schematic mentions; "measure at 117VAC". But I'm sort of curious if maybe they measured at 115VAC? The Sam's schematic falls more in line with voltages stated measured at 117VAC from what I come up with. If a guy looks at the values for both Pilot and Sam's schematics, and determine total plate current from cathode voltage, they both just hammer on the 6BQ5s, if my math is right... You mention the amplifier (sounds?), or (operates?) best with the tubes at near max dissipation? I can just imagine back in 1960... "This amplifier sounds wonderful for a few months, and then the tubes glow red." Seems kinda rich even back when readily available 6BQ5/EL84 were 3-4 bucks a pop. Using a combination variac/isolation transformer, with 117VAC in, I get 360 volts B+ on the first cap, with 335 or so on the plates, and 12 volts on the cathodes. My numbers for dissipation of the output tubes are the same as you mentioned. I swapped in 150 cathode resistors, it raised the plate voltage 3-4 volts, and lowered the cathode voltage about a volt. It still comes out around 13-14 watts per tube. Keep in mind I'm trying to average things out, because I'm not using matched output tubes. I don't really want to cook my good 6BQ5s, so I'm using a junk quad to get going. They match emission-wise on a tester, but that's pointless in circuit. If I sub in a 5R4 rectifier, I get 312 on the first cap, around 285-290 on the plates, and around 11 volts on the cathodes. The drops the dissipation number down to 10 watts for each tube. Roughly 85% of the total 12-watt dissipation rating I'm guessing. It doesn't drop much voltage on the driver and phase inverter tubes, so it may actually pass a signal. The reason why I said these amps run the tubes near max dissipation is because you pretty much have to with cathode bias or else when they go into Class B you will get some nasty crossover distortion. Ideally you want to bias these amps for around 11-12 watts plate dissipation, the tubes will last longer and you don't get the crossover distortion. It's when these amps push the tubes well past 12 watts is when the tubes crap out once a year, it's just too much for them. Also tubes back then lasted longer and took more abuse for the most part. I see people chasing their tails upping the cathode resistor because with less current the B+ increases and you kinda end up back to where you were in regard to dissipation. Except now you have pushed the operating point over to where you will get crossover distortion so it's a lose lose situation, you don't really reduce dissipation but you lower the current changing the operating point making it sound awful from the crossover distortion. The only way to fix these amplifiers is to reduce the B+ to the output transformer. I still like to use 150 ohms, with 11 volts at cathode your at 37mA per tube which puts your target plate voltage around 335v for 12 watts dissipation. 335 - 11 = 324 11/150 = .073 .073 / 2 = .037 or 37mA 12 watts / .037 = 324 volts So when I reduced the B+ I had a goal of getting ~335v on the plates. Quote Link to comment Share on other sites More sharing options...
captainbeefheart Posted January 4, 2022 Share Posted January 4, 2022 Don't forget the cathode voltage across the cathode resistor is derived from the sum of plate current and screen current, So if you aim for 12 watts dissipation your really at around 11 watts. So even if you are getting something like 12v on the cathode with the 150 ohm cathode resistor and you calculate out 13 watts, don't panic because your really not quite there because of the screen current. You can of course measure the screen current and deduct it from the combined cathode current you calculate and get a more precise dissipation but as long as you understand this concept I don't bother, I just make sure I am 13 watts or under on the calculations and I can sleep sound at night knowing the tubes are within their maximum ratings. I have had tubes running like this for years, it's when things go unchecked and your at 120% max dissipation or more the tubes just burn out quick. Quote Link to comment Share on other sites More sharing options...
mike stehr Posted January 4, 2022 Author Share Posted January 4, 2022 I went ahead and checked out both channels on the oscilloscope using the 5R4 rectifier. Before I connected the signal generator, I went ahead to see how the output channel A and B looked on the scope without a signal. With the inputs shorted. Channel A has a nice noisy looking ripple waveform going. Channel B has the same ripple waveform with a nice spike in the middle...almost like a square wave sort of spike with a sharp slope. Too bad I can't recall if the slope was leading or lagging...dur... I didn't bother to connect the signal generator. It would be too difficult to see the waveform through the ripple. 5 hours ago, captainbeefheart said: The only way to fix these amplifiers is to reduce the B+ to the output transformer. Indeed. Even my local audio mentor brought it up. To be honest, I have no problems with the Pilot schematic/circuit. But I am not fond of how the circuit is laid out in too small of a chassis, with a weird Pi filter that parks B+ for the OPTs at the screen pin of one output tube of the push-pull pair. I measure the screen pins, and I have more voltage on the screens than the plates. (The two tubes without the 270-ohm screen resistors) Or the fact everything is grounded to the sockets, and many chassis' grounds as well. I suppose I could connect the amplifier to my bench speakers to hear how noisy it is or figure out why. But it this point, I'm really pondering buying a bud/hammond aluminum chassis of larger dimensions. Then I would have room to lay-out/wire the circuit my way. With a choke loaded supply, or CLC to drop B+ with a bit better regulation. Before I do that, I best make sure these output transformers are up to snuff...I don't like that spike... I'll determine the primary impedance and run a simulated primary resistance with a resistor and an 8-ohm load resistor. Then run a signal generator into the primary and monitor the 8-ohm secondary with a scope using sine and square waves. Monitor for any anomalies between each output transformer, and if the characteristics match up for both pair. Thanks for the help, Mike Quote Link to comment Share on other sites More sharing options...
John Warren Posted January 4, 2022 Share Posted January 4, 2022 10 hours ago, captainbeefheart said: You can of course measure the screen current 100_UL_Screen_Grid_Dissipation.pdf Quote Link to comment Share on other sites More sharing options...
captainbeefheart Posted January 4, 2022 Share Posted January 4, 2022 What frequency and amplitude is the ripple at the output? That may help us track the source down. There is a voltage divider coming off the 225v node, 390k and 39k 39000 / (39000+ 3900000) = .09 .09 * 225 = 20v So the heater center tap is elevated to 20vdc Quote Link to comment Share on other sites More sharing options...
Curious_George Posted January 6, 2022 Share Posted January 6, 2022 I have a Realistic SAF-24 (circa 1961, made in Japan) that is similar to the Pilot SA-232. I've been meaning to restore it for years, but have not yet. Last time I plugged it in it worked great. Quote Link to comment Share on other sites More sharing options...
Curious_George Posted January 6, 2022 Share Posted January 6, 2022 Not sure what RIAA curve it had when made, but I now remember redesigning the phono section for the "contemporary" RIAA curve that is accepted today as "standard". Quote Link to comment Share on other sites More sharing options...
erik2A3 Posted January 15, 2022 Share Posted January 15, 2022 Plate dissipation and triangle waves aside for a moment, something as simple as checking (and correcting as needed) filament voltages on vintage equipment originally designed for lower mains supply is also important. If you’re planning on using a variac long-term, not an issue, as you of course know. I did the same work on a LEAK ST20 ten years or so ago. Push-pull or single ended, EL84s are really good. Easily the equal of some of the outrageously priced output triodes. As always, your workmanship and attention to detail are among the very best we see here. Very nicely done. Quote Link to comment Share on other sites More sharing options...
erik2A3 Posted January 15, 2022 Share Posted January 15, 2022 You might also achieve a little HT sag with a different rectifier (obviously with suitable current specs). I’ve also had good results using a thermistor between mains input and PT primary. Great project, and I have to thank you and other contributors here for inspiring me to get back to my workbench. Quote Link to comment Share on other sites More sharing options...
sunnysal Posted January 25, 2022 Share Posted January 25, 2022 Great work Mike! Please keep us updated as to progress. I still am waiting for an excellent el84 amp to pass in front of me so I can buy it. Since I have sworn off DIY (or better said my wife has made me swear it off, at least the high voltage stuff). I may end up buying a new one like; Roger M's RM-10 1 Quote Link to comment Share on other sites More sharing options...
mike stehr Posted February 6, 2022 Author Share Posted February 6, 2022 A different approach... 1 Quote Link to comment Share on other sites More sharing options...
mike stehr Posted February 19, 2022 Author Share Posted February 19, 2022 A few more holes... Quote Link to comment Share on other sites More sharing options...
henry4841 Posted February 20, 2022 Share Posted February 20, 2022 How are you making holes? I just use a step drill bit myself, sometimes referred to a Christmas tree bit. Not nearly as good as a punch, those things are expensive, but good enough with some dressing the rough edges. Quote Link to comment Share on other sites More sharing options...
mike stehr Posted February 20, 2022 Author Share Posted February 20, 2022 4 hours ago, henry4841 said: How are you making holes? I just use a step drill bit myself, sometimes referred to a Christmas tree bit. I used a stepped bit. I have a smaller one, and a larger one that maxes out at 1-3/8". How often will one need a 1-3/8" chassis punch/die? They can leave a nasty burr on the bottom of the piece being drilled. But if one can flip the piece over and get at the burred side, one can use the bit to de-burr the hole. I think the key is to keep the piece being drilled as anchored as possible along with the press. (I have a cheap tabletop) The bits are aggressive, especially larger one. Quote Link to comment Share on other sites More sharing options...
captainbeefheart Posted February 20, 2022 Share Posted February 20, 2022 I hate drilling so I draw up what I need in a CAD program and bring it to a machine shop to make. It's a little bit more money but not much, it's less work I have to do and the results end up perfect. 2 Quote Link to comment Share on other sites More sharing options...
mike stehr Posted February 24, 2022 Author Share Posted February 24, 2022 1 Quote Link to comment Share on other sites More sharing options...
Curious_George Posted February 24, 2022 Share Posted February 24, 2022 On 2/20/2022 at 7:56 AM, henry4841 said: How are you making holes? I just use a step drill bit myself, sometimes referred to a Christmas tree bit. Not nearly as good as a punch, those things are expensive, but good enough with some dressing the rough edges. I've been using a punch set for about 20 years. At first, the set may seem expensive, but the final cosmetic results will be worth the price of admission. It makes perfectly round holes and zero burrs. I also use a "Unibit" for other small holes typically 3/4" and less. The punch set has worked well with just about all tube sockets, but you do have to be selective of the socket(s). A little creative thinking will usually get you past any roadblocks you encounter with the tube socket hole being the correct size. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.