Concerning #85, that schemo makes no sense. I strongly suspect that what you'd get there is an oscillator. The grids of the upper triodes of the cascodes needs to be returned to AC ground, not the cathodes. The cascode is a two stage amp: grounded cathode --> grounded grid cascade. A grounded grid amp goes unstable if the grid-to-ground return has too much impedance, and this one doesn't even return the grids to ground. 
Not. A. Good. Idea.
Not. A. Good. Idea.
Attachments
SY said:
Got any examples? I've never seen that done with SS cascodes, or hollow state ones either till that particular post.
The explanation earlier in the thread made no sense at all. Yeah, they do sometimes consider the series connection to be a "composite tube", with a sort of tetrode-like characteristic sans the secondary emmission kinks. (I've never found that to be useful, and design cascodes as a two stage amp.) If you float the cathode of a tetrode, you'd want to couple the screen to the cathode so's to keep the screen-to-cathode voltage constant. However, this isn't a tetrode. It's a two stage amp, and if you eliminate the grounded grid stage grid-to-cathode voltage, you eliminate the signal and the output. After all, you can AC couple a grounded cathode stage to a grounded grid stage to get a legitimate circuit called a "folded cascode". It's still a two stage amp if DC coupled.
If you don't ground the grid of a grounded grid stage, it will oscillate. After all, those UHF triodes that operate at 400MHz+ have as many as five connections to the grid (e.g. 6AM4) so that paralleling them to the ground plane minimizes the grid-to-ground impedance. That internal base impedance (r(b) ~= 20R for Si BJTs) is why the grounded base topology doesn't have the unconditional stability of the grounded grid amp.
When looking at (2) of the tubes available on different driver boards the 7199 and the 6GH8 which is better?
One site selling a driver board claims the 7199 is superior. One site pushing the 6GH8 claims it to be superior.
Anyone conducted distortion tests that can finally put an end to this question?
One site selling a driver board claims the 7199 is superior. One site pushing the 6GH8 claims it to be superior.
Anyone conducted distortion tests that can finally put an end to this question?
When looking at (2) of the tubes available on different driver boards the 7199 and the 6GH8 which is better?
One site selling a driver board claims the 7199 is superior. One site pushing the 6GH8 claims it to be superior.
Anyone conducted distortion tests that can finally put an end to this question?
One site selling a driver board claims the 7199 is superior. One site pushing the 6GH8 claims it to be superior.
Anyone conducted distortion tests that can finally put an end to this question?
Heh, heh, heh.
Yes, I have. The distortion of both is all over the map with neither tube looking definitively superior. There's really not much to choose between one and the othe, but there are some individual tubes within each type that are better and some that are worse.
MUCH more variation than, say, among small signal triodes. Much higher distortion, too, but you'd expect that.
Yes, I have. The distortion of both is all over the map with neither tube looking definitively superior. There's really not much to choose between one and the othe, but there are some individual tubes within each type that are better and some that are worse.
MUCH more variation than, say, among small signal triodes. Much higher distortion, too, but you'd expect that.
hey-Hey!!!,
With any of the small signal pentodes, the g2 voltage/load match is quite important. Screw that up, and it won't matter what pentode you get, performance will suffer. Either way, I'd go for single-element tubes to get the specs I want. I suspect a 6EJ7 would do for the pentode, and there's *MANY* single triodes to look at.
cheers,
Douglas
With any of the small signal pentodes, the g2 voltage/load match is quite important. Screw that up, and it won't matter what pentode you get, performance will suffer. Either way, I'd go for single-element tubes to get the specs I want. I suspect a 6EJ7 would do for the pentode, and there's *MANY* single triodes to look at.
cheers,
Douglas
burnedfingers said:
back 'in the day' (before the internet), new 7199s were hard to find. I modified my own '70 and a pair of Mark IVs to run 6GH8s and I couldn't hear any earth-shattering difference.
These days I wouldn't bother with either the 7199 or 6GH8 - there are plenty of other drop-in boards worth checking out - I'm running the Poseidon boards (5751 driver /12BH7 LTP) in a pair of Mark IIIs and they work great (nicely made board too).
7199's
The 7199's were an excellent sounding tube from the right manufacturer. I always sought out Phillips ECG's, and when they got rare- Sylvania's, and always in matched pairs. I never cared for anyone elses. I gave my beloved (and modded)ST-70 to my son for his birthday when I started my DIY ST-70 project, which was to use 3)6201's. I Got ripped by 'Handwound Transformers on the transformer set(AKA 'Lucas Electronics'- see part of the story on this site somewhere), got disgusted(shame on me) and gave up the project. I just listed the rest of my parts and completed VTA board on eBay, if anyone's interested(same username as in here).
The 7199's were an excellent sounding tube from the right manufacturer. I always sought out Phillips ECG's, and when they got rare- Sylvania's, and always in matched pairs. I never cared for anyone elses. I gave my beloved (and modded)ST-70 to my son for his birthday when I started my DIY ST-70 project, which was to use 3)6201's. I Got ripped by 'Handwound Transformers on the transformer set(AKA 'Lucas Electronics'- see part of the story on this site somewhere), got disgusted(shame on me) and gave up the project. I just listed the rest of my parts and completed VTA board on eBay, if anyone's interested(same username as in here).
Its nice to be able to use the 7199 or do a slight modification to the board and pop in a 6GH8.
The distortion figures according to SY are running neck and neck on the two so nothing to gain either way.
I've had the VTA board and while it gave the ability to adjust individual tube bias the board was nothing special. I didn't hear any earth shattering differences between it and the stock board and the same goes for the Poseiden and some like it. The biggest difference is to be able to pick different tubes and do a little rolling.
I actually have heard the biggest differences in the stock board by updating both the capacitors and replacing all the resistors with new hand picked 1% pieces.
The distortion figures according to SY are running neck and neck on the two so nothing to gain either way.
I've had the VTA board and while it gave the ability to adjust individual tube bias the board was nothing special. I didn't hear any earth shattering differences between it and the stock board and the same goes for the Poseiden and some like it. The biggest difference is to be able to pick different tubes and do a little rolling.
I actually have heard the biggest differences in the stock board by updating both the capacitors and replacing all the resistors with new hand picked 1% pieces.
I absolutely agree that the most audible gain on the stock board can be had by swapping out those antiquated Caps and resistors. I went to 1% matched MIT MultiCaps and carbon film resistors all through the signal path back in '97, slightly beefed up the PS, rewired the interior, bypassed the mono switch, and installed a much better cord, Vampire gold RCAs and 5-ways(all without chassis butchery). I thought about 6GH8s, but never thought they sounded nearly as musical as the 7199's. There are a multitude of excellent 6201 family triodes out there to use though. With the extra B+ available(6 times that of stock) with the VTA mod: they CAN make a dramatic improvement over the original. I've still got the four(figured a spare was a good idea) Telefunken ECC801S tubes I bought for my project. I'm still kicking myself in the butt for not finishing it, and hoping Dave(Handwound Transformers) Lucas' armpits have been infested by the fleas of a thousand camels.
Dynaco ST-70 transformer heat
Hi Guys,
I just got my ST-70 back up and running after a long time sitting in the closet (this is the second time it has been restored) and I have a question about the power transformer temperature.
I had it up and running yesterday, and all the voltages are perfect, bias is set and stable, and I let it run for 8 hours. Now I know the power transformers run hot, but I'm not sure how hot is safe. I got the digital thermometer out and propped it in one of the screw holes and left it for 30 minutes, which gave me a reading of 75 degrees C.
Is this an acceptable temp for this thing, or is it on its way out? There's a significant level of hum from it, which actually induces some mechanical hum into the output by vibrating the driver tubes. I know this because the hum stops when I squeeze the chassis or the tubes.
I guess I could pick up one of the new ones from Triode Electronics, but I'm not too keen on dropping another $140 on this thing, as it has already bled me dry. Twice...
I've attached a picture of the finished product, so feel free to let me know what you think. Just as a pertinent note, the heaters on the damper diodes are on a separate transformer, so the main power transformer is under even lighter load than usual.
Cheers,
Owen
Hi Guys,
I just got my ST-70 back up and running after a long time sitting in the closet (this is the second time it has been restored) and I have a question about the power transformer temperature.
I had it up and running yesterday, and all the voltages are perfect, bias is set and stable, and I let it run for 8 hours. Now I know the power transformers run hot, but I'm not sure how hot is safe. I got the digital thermometer out and propped it in one of the screw holes and left it for 30 minutes, which gave me a reading of 75 degrees C.
Is this an acceptable temp for this thing, or is it on its way out? There's a significant level of hum from it, which actually induces some mechanical hum into the output by vibrating the driver tubes. I know this because the hum stops when I squeeze the chassis or the tubes.
I guess I could pick up one of the new ones from Triode Electronics, but I'm not too keen on dropping another $140 on this thing, as it has already bled me dry. Twice...
I've attached a picture of the finished product, so feel free to let me know what you think. Just as a pertinent note, the heaters on the damper diodes are on a separate transformer, so the main power transformer is under even lighter load than usual.
Cheers,
Owen
Attachments
Hi Owen,
75C is pretty hot, but might be ok depending on whether or not the actual winding temperature is not too much hotter. You can use the rise of resistance measurement technique and measure the winding resistance cold and hot and calculate the actual winding temperature from that. I would use the high voltage secondary as it has fairly high resistance, and is likely the hottest part of the transformer.
Measure the entire winding resistance directly at the rectifier tube socket(s) and be sure to get the most accurate resistance measurement possible. Measure the meter lead resistance and subtract that from the measured value. Obviously a 4 wire resistance measurement (kelvin connected) would be best, but unless you have a lab grade meter this won't be possible.
Needless to say: UNPLUGGED, and caps safely discharged.
Equation: Delta T = (Rhot - Rcold)/(Rcold x 0.003931)
Copper tc is 3931ppm per degree C.. FYI
You need to know the ambient temperature at the start of the test and you add this value to Delta T to determine the actual winding temperature. Should it be over 100C I would be very concerned.
A small fan blowing on the amplifier will solve this problem.
75C is pretty hot, but might be ok depending on whether or not the actual winding temperature is not too much hotter. You can use the rise of resistance measurement technique and measure the winding resistance cold and hot and calculate the actual winding temperature from that. I would use the high voltage secondary as it has fairly high resistance, and is likely the hottest part of the transformer.
Measure the entire winding resistance directly at the rectifier tube socket(s) and be sure to get the most accurate resistance measurement possible. Measure the meter lead resistance and subtract that from the measured value. Obviously a 4 wire resistance measurement (kelvin connected) would be best, but unless you have a lab grade meter this won't be possible.
Needless to say: UNPLUGGED, and caps safely discharged. Equation: Delta T = (Rhot - Rcold)/(Rcold x 0.003931)
Copper tc is 3931ppm per degree C.. FYI
You need to know the ambient temperature at the start of the test and you add this value to Delta T to determine the actual winding temperature. Should it be over 100C I would be very concerned.
A small fan blowing on the amplifier will solve this problem.
You also have the "cost-reduced" version of the PA-060. The original ones had a taller core stack and were not potted in black goop. The smaller ones did run hotter... and I never liked them. Also note that the original design pretty much ran the power transformer at it's full rated output. Your rebuild appears to show two rectifiers instead of one.... have you calculated the total current draw with the new tube compliment?
Regards, KM
Regards, KM
kmaier said:
Quite true.. I noticed the same thing. I hope you are using a separate filament transformer for what appears to be a pair of damper diodes.
The existing filament windings can run a 2 tube external pre-amp per channel or a couple of extra small signal tubes in a modded version, but not the several extra amps of damper diode filaments.
Hot Transformer
Hey Guys,
I mentioned the filament supply in the last line of my post, but I should have put it at the top. I'm running a separate 12V 2A transformer under the chassis feeding the two 6DE4 damper diodes in series, so they're not loading the original transformer.
The new front end (12AU7 and EF86 x 2) actually draws about 0.2 amps less than the original 7199's so that's not a problem either.
Technically, the transformer has about 15 watts less load than it normally would, but it's still cooking.
Measuring the winding resistance is a great idea, and I'm in the process now. I took the hot measurement and it was exactly 104.9 ohms about two seconds after I flicked it off and pulled the damper diodes out. I'm now waiting for it to get down to ambient temp before taking the second resistance measurement.
They should change that saying to "A watched transformer never cools"
Hey Guys,
I mentioned the filament supply in the last line of my post, but I should have put it at the top. I'm running a separate 12V 2A transformer under the chassis feeding the two 6DE4 damper diodes in series, so they're not loading the original transformer.
The new front end (12AU7 and EF86 x 2) actually draws about 0.2 amps less than the original 7199's so that's not a problem either.
Technically, the transformer has about 15 watts less load than it normally would, but it's still cooking.
Measuring the winding resistance is a great idea, and I'm in the process now. I took the hot measurement and it was exactly 104.9 ohms about two seconds after I flicked it off and pulled the damper diodes out. I'm now waiting for it to get down to ambient temp before taking the second resistance measurement.
They should change that saying to "A watched transformer never cools"
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.