whoa! what happened to the 125ESE? 😱 don't you have a source for end bells? sorry, it's just me... 
Yeah, I decided to just not bother with taking the transformers apart to add bell ends. I thought about putting the 125ESE's under the hood, but it's getting a little crowded in there already.
Hi Saurav;
That amp's shaping up fine!
>>>...This is definitely my weakest area, painting metal. I always get blotches on there with spray paint, and sometimes the paint doesn't stick...<<<
One question about the 'sticking problem': did you clean and thoroughly degrease the metal before priming? If not, on your next project you might try something like acetone - and DO NOT touch the chassis with your hands after cleaning it (wear clean gloves or handle the chassis only from the underside). Oils from your skin can cause all sorts of problems with finishes....
The best way I've found to get a good paint finish is to work with several thin coats rather than a single thick one. As far as the blotches are concerned, you can GENTLY sand them down smooth and repaint over top of them. Start with a 400 grit for big nasty runs or brush marks in the paint, 600 grit as the second step, and 1500 grit for the final finishing. During the later steps, you should be very VERY gentle with the sanding. You can get 600 and 1500 grit papers at any auto shop. After you've got a mirror gloss going for you, you can apply a transparent topcoat to help protect the finish.
Good luck and all the best,
Morse
That amp's shaping up fine!
>>>...This is definitely my weakest area, painting metal. I always get blotches on there with spray paint, and sometimes the paint doesn't stick...<<<
One question about the 'sticking problem': did you clean and thoroughly degrease the metal before priming? If not, on your next project you might try something like acetone - and DO NOT touch the chassis with your hands after cleaning it (wear clean gloves or handle the chassis only from the underside). Oils from your skin can cause all sorts of problems with finishes....
The best way I've found to get a good paint finish is to work with several thin coats rather than a single thick one. As far as the blotches are concerned, you can GENTLY sand them down smooth and repaint over top of them. Start with a 400 grit for big nasty runs or brush marks in the paint, 600 grit as the second step, and 1500 grit for the final finishing. During the later steps, you should be very VERY gentle with the sanding. You can get 600 and 1500 grit papers at any auto shop. After you've got a mirror gloss going for you, you can apply a transparent topcoat to help protect the finish.
Good luck and all the best,
Morse
Just looked at your chassis picture, and I'm wondering if those ceramic 4-pin sockets have any clearance under the chassis. The rivets are connected to the terminals, and you want 1/8" (or an insulator) berween them and the chassis.
Yup. I'd tried vinegar in the past without much success, so this time I actually bought "metal etch". The primer's bonded to the metal fine, when the paint peels off I see primer underneath. Maybe I have the wrong kind of primer.
Next time I'll try sanding. I wasn't sure if I could do that.
The rivets are a little countersunk though, and are below the surface of the ceramic. Probably not 1/8" though. In any case, I'll double check, thanks for the warning. What kind of insulator could I use? It would have to handle fairly high temperatures, I would think?
Well, I didn't post this last night. I completed it, nothing blew up or caught on fire, and it sounds fabulous. It's the old "I'm hearing things I never heard before" cliche. The music sounds smooth and relaxed, with the little details being more easily audible that before.
I still need to fix a few things. I totally miscalculated the amount of current the 6SL7s would draw, so the B+ on the driver tubes is way too high, I need to buy larger value resistors today. And the wiring's somewhat messy because I decided to leave all transformer leads at their original lengths. I managed to tuck everything away near the rear of the chassis where the PS is, so I hope that won't affect the sound too much. It sounds pretty quiet with just an adjustment of the hum pots. With 97dB speakers, I can barely hear the hum 2' away.
Thanks for all the help, everyone.
I still need to fix a few things. I totally miscalculated the amount of current the 6SL7s would draw, so the B+ on the driver tubes is way too high, I need to buy larger value resistors today. And the wiring's somewhat messy because I decided to leave all transformer leads at their original lengths. I managed to tuck everything away near the rear of the chassis where the PS is, so I hope that won't affect the sound too much. It sounds pretty quiet with just an adjustment of the hum pots. With 97dB speakers, I can barely hear the hum 2' away.
Thanks for all the help, everyone.
Hi Saurav!
Great!! Glad to read that your amp's working well and that you're happy!! Isn't it a magical moment, when your creation starts to produce music for the first time? It's addicting though, so you'll soon start thinking of building another - maybe a litlle one for the den or the office..... ;-)
>>>...The primer's bonded to the metal fine, when the paint peels off I see primer underneath. Maybe I have the wrong kind of primer...<<<
Hmmmm, you've got me stumped. I've had troubles in the past with getting things to stick to the metal but not the primer. What kind of paint did you use? I usually use auto body type enamels, though some of my friends who used to build telescopes swear by epoxy paints (they'd use 'em on the tubes and mounts, which can get scraped, etc in handling).
Good luck with your amp and all your projects!!
All the best,
Morse
Great!! Glad to read that your amp's working well and that you're happy!! Isn't it a magical moment, when your creation starts to produce music for the first time? It's addicting though, so you'll soon start thinking of building another - maybe a litlle one for the den or the office..... ;-)
>>>...The primer's bonded to the metal fine, when the paint peels off I see primer underneath. Maybe I have the wrong kind of primer...<<<
Hmmmm, you've got me stumped. I've had troubles in the past with getting things to stick to the metal but not the primer. What kind of paint did you use? I usually use auto body type enamels, though some of my friends who used to build telescopes swear by epoxy paints (they'd use 'em on the tubes and mounts, which can get scraped, etc in handling).
Good luck with your amp and all your projects!!
All the best,
Morse
Rust-Oleum Metallic Copper, and as far as I remember the primer was Rust-Oleum too. I'm not sure what medium the paint is.
OK, more questions.
How hot should a 10W sandcast resistor get? I put in a pair of 100 ohm 10W resistors in series between the rectifier and the first filter cap/choke, to drop the B+ voltage a bit. Here's my calculation:
RMS current draw from rectifier: ~ 120 - 150mA. Figure 60mA for both 2A3s, about 1.5mA for each 6SL7, 150mA is a good highish figure. For each 100 ohm resistor, that's an RMS power dissipation of 1.5 - 2.5W per resistor. Now these resistors aren't seeing the RMS current draw, they're seeing the current surges as the cap charges during the duty cycle. If I assume a duty cycle of 0.1, that's 1.5A surges, which puts me at 225W dissipation.
Now 225W doesn't seem like a reasonable value to use for the required power rating in this position. Is 10W enough, or should I get higher rated resistors? I ran the amp for a minute or so to measure bias voltages, and these resistors got hot enough that I wouldn't want to touch them for more than a couple of seconds. And based on my voltage measurements, my estimates for the current draw are fairly close.
Or should I re-model the PSU with a lower valued initial cap? Won't adding dropping resistors increase the output impedance of the power supply, and cause it to sag more on transients?
Thanks in advance,
Saurav
How hot should a 10W sandcast resistor get? I put in a pair of 100 ohm 10W resistors in series between the rectifier and the first filter cap/choke, to drop the B+ voltage a bit. Here's my calculation:
RMS current draw from rectifier: ~ 120 - 150mA. Figure 60mA for both 2A3s, about 1.5mA for each 6SL7, 150mA is a good highish figure. For each 100 ohm resistor, that's an RMS power dissipation of 1.5 - 2.5W per resistor. Now these resistors aren't seeing the RMS current draw, they're seeing the current surges as the cap charges during the duty cycle. If I assume a duty cycle of 0.1, that's 1.5A surges, which puts me at 225W dissipation.
Now 225W doesn't seem like a reasonable value to use for the required power rating in this position. Is 10W enough, or should I get higher rated resistors? I ran the amp for a minute or so to measure bias voltages, and these resistors got hot enough that I wouldn't want to touch them for more than a couple of seconds. And based on my voltage measurements, my estimates for the current draw are fairly close.
Or should I re-model the PSU with a lower valued initial cap? Won't adding dropping resistors increase the output impedance of the power supply, and cause it to sag more on transients?
Thanks in advance,
Saurav
And pictures:
That's a red lamp shining on the amp, not its natural color. I added a .022 cap from the input RCA to the grid of the first 6SL7 to roll off the bass going to the 2A3s, 125ESEs and most importantly my speakers. You can also see the extra 125ESE leads just tie-wrapped together in the middle. Any advice about a better way to handle these wires would be appreciated. However, my biggest worry right now are the PS dropping resistors that I mentioned in the previous post.
Thanks,
Saurav
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
That's a red lamp shining on the amp, not its natural color. I added a .022 cap from the input RCA to the grid of the first 6SL7 to roll off the bass going to the 2A3s, 125ESEs and most importantly my speakers. You can also see the extra 125ESE leads just tie-wrapped together in the middle. Any advice about a better way to handle these wires would be appreciated. However, my biggest worry right now are the PS dropping resistors that I mentioned in the previous post.
Thanks,
Saurav
One more question. Pin 1 on a 5AR4 is usually shown as a dotted line, with N.C/I.C. written on it. I assume that means not connected or internall connected. However, there are other pin positions that aren't used on this tube (and the pins aren't even present), so what is special about pin 1? Is it a shield that I'm supposed to connect to ground? Or should I just ignore it?
N.C. or I.C. translates as "don't connect to this pin!" Often, it's a wire going up to help support a heavy structure like an anode, but it can vary between manufacturers, so one valve might use the pin and another not use it.
Thanks. Could you please help me out with my question regarding B+ dropping resistors? It's about 6 posts up from here 🙂 I think I made 4 posts in a row to this thread.
I`m not an expert on valve amps but I think if your using a resistor to drop your voltage then that 10 watt resistor is being overworked a wee bit from what you have said....there are more experienced members here with tube audio p/s`s that can assist you further...just be patiant😉
DIRT®
DIRT®
If you can't stand the heat...
Those resistors will get hot, which is fine for them, but not so good for anything nearby.
If you are using PSUD for modelling your power supply, you will find that it can give an RMS current (in the table on LHS) for this resistor which you can use for your heating calculation. I doubt if you are disspiating more than 2.5W, but where is the heat going to go? When you use your amplifier, the resistor is enclosed by the chassis and there isn't a free flow of air, so it heats up the internal air until the air heats the chassis, which can then lose heat to the surrounding air. I see you've used metal-clad resistors bolted down to the chassis elsewhere, I'd use some more for the 100R. (And don't forget a thin smear of heatsink compound before fitting the resistor.)
Those resistors will get hot, which is fine for them, but not so good for anything nearby.
If you are using PSUD for modelling your power supply, you will find that it can give an RMS current (in the table on LHS) for this resistor which you can use for your heating calculation. I doubt if you are disspiating more than 2.5W, but where is the heat going to go? When you use your amplifier, the resistor is enclosed by the chassis and there isn't a free flow of air, so it heats up the internal air until the air heats the chassis, which can then lose heat to the surrounding air. I see you've used metal-clad resistors bolted down to the chassis elsewhere, I'd use some more for the 100R. (And don't forget a thin smear of heatsink compound before fitting the resistor.)
Joe,
Thanks for the encouragement. I took the resistors out for tonight. I figured I'm better off running my tubes a little hot and shortening their life a little, than having my amp catch on fire 🙂
EC8010,
I thought that with resistors placed before the first filter cap, RMS current didn't apply because the current through the resistor is in pulses, not a steady draw?
The metal-clad resistors are for the 2A3's cathode bias. I saw some that looked like that at the surplus store today, I'll go back there tomorrow and see if I can find something suitable. The heatsink compound is a good idea.
In general, is it better to lower B+ by changing capacitor values? Seems to me that adding resistors would make the PS output voltage sag more during high current demand. Or is that not an issue with a class A SE amp?
Thanks for the encouragement. I took the resistors out for tonight. I figured I'm better off running my tubes a little hot and shortening their life a little, than having my amp catch on fire 🙂
EC8010,
I thought that with resistors placed before the first filter cap, RMS current didn't apply because the current through the resistor is in pulses, not a steady draw?
The metal-clad resistors are for the 2A3's cathode bias. I saw some that looked like that at the surplus store today, I'll go back there tomorrow and see if I can find something suitable. The heatsink compound is a good idea.
In general, is it better to lower B+ by changing capacitor values? Seems to me that adding resistors would make the PS output voltage sag more during high current demand. Or is that not an issue with a class A SE amp?
RMS current always applies - it's just that it's usually hard to calculate, and expensive to measure (have a look at the price of proper true-RMS meters). RMS is an equivalent DC current that has the same heating effect as any pulse or AC waveform.
SE amplifiers are more critical of their power supply than PP. Ideally, they would like a regulated HT. Either series resistors or tweaking reservoir capacitor value allow you to drop HT volts. Tweaking reservoir value is slightly better because it doesn't waste power (increasing heat within the chassis). Both methods make the power supply sag more under load.
SE amplifiers are more critical of their power supply than PP. Ideally, they would like a regulated HT. Either series resistors or tweaking reservoir capacitor value allow you to drop HT volts. Tweaking reservoir value is slightly better because it doesn't waste power (increasing heat within the chassis). Both methods make the power supply sag more under load.
RMS:
As EC8010 says, RMS always applies. It's the area under the graph line that determines it. The 0.707*peak rule only works for sine-waves, hence his recommendation to use PSUD to do the work for you.
Placement of hot resistors:
Remember that the normal power rating (w) is only usable at or just above ambient. Put them in a hot chassis and they must be de-rated. See attached picture from:
http://www.irctt.com/pdf_files/PWRL.pdf
Keep them away from electrolytics and other heat sensitive components.
As EC8010 says, RMS always applies. It's the area under the graph line that determines it. The 0.707*peak rule only works for sine-waves, hence his recommendation to use PSUD to do the work for you.
Placement of hot resistors:
Remember that the normal power rating (w) is only usable at or just above ambient. Put them in a hot chassis and they must be de-rated. See attached picture from:
http://www.irctt.com/pdf_files/PWRL.pdf
Keep them away from electrolytics and other heat sensitive components.
Attachments
OK, I think I understand now. I can't just add up the RMS currents of the tubes to get the RMS current through the PS resistor, or use RMS equations that apply to sine waves. But PSUD will calculate the RMS current for me, so I can use that value for the power calculations. And PSUD says the RMS current is ~ 200mA, which is more than (2x60 + a little bit), so that makes sense now.
Didn't know that.
Thanks a lot, this is a lot clearer to me now.
Didn't know that.
Thanks a lot, this is a lot clearer to me now.
Power supply sag.
It is true that SE is power supply critical, this being because the power supply is in series with the output signal. But being class A, the current average demand does not theoretically change with signal. This is true only as long as the stage introduces no distortion. So in practice it means that there is hardly any change at normal signal levels, but a significant change when clipping is reached.
If you do not have access to a distortion meter, you can actually use the voltage change to determine the distortion.
BTW Nice work so far 😉
It is true that SE is power supply critical, this being because the power supply is in series with the output signal. But being class A, the current average demand does not theoretically change with signal. This is true only as long as the stage introduces no distortion. So in practice it means that there is hardly any change at normal signal levels, but a significant change when clipping is reached.
If you do not have access to a distortion meter, you can actually use the voltage change to determine the distortion.
BTW Nice work so far 😉
- Status
- Not open for further replies.