For an OTL Power Amplifier with 4 6C33C, i'm using an ATX Computer PSU for the heaters. I'm running each 6C33C at 12V@3.3A which means for four units, they'll consume approx 160W, a number i believe my PSU is capable of supplying.
One problem: cold starts. With 4 6C33C turned on at the same time, the ATX PSU just refuses to start. I believe it's because the heaters are cold and have very low resistance, causing surge to the PSU and then the PSU reads them as short circuits. I can plug just one 6C33C and the PSU will turn on and supply 12V normally.. but once i plug another one, same problem occurs.
I'm thinking of connecting the heaters in series with some resistors for initial startup. Then maybe after 1-2 minutes, i will bypass the resistor for full heater power. Does this sound feasible? What should the resistance value for each 6C33C? Perhaps 100R/5W --> meaning at inital starts, at worse there is only 12V/100R = 120mA flowing for each tube? I'm still not sure if the PSU will continue to work if i then bypass the resistors one at a time.
One problem: cold starts. With 4 6C33C turned on at the same time, the ATX PSU just refuses to start. I believe it's because the heaters are cold and have very low resistance, causing surge to the PSU and then the PSU reads them as short circuits. I can plug just one 6C33C and the PSU will turn on and supply 12V normally.. but once i plug another one, same problem occurs.
I'm thinking of connecting the heaters in series with some resistors for initial startup. Then maybe after 1-2 minutes, i will bypass the resistor for full heater power. Does this sound feasible? What should the resistance value for each 6C33C? Perhaps 100R/5W --> meaning at inital starts, at worse there is only 12V/100R = 120mA flowing for each tube? I'm still not sure if the PSU will continue to work if i then bypass the resistors one at a time.
Are you sure your PSU is able to deliver 13 Amps on its 12V output alone ?
Anyway . . .
Transformateurs - Transformateurs toriques - VELLEMAN 16012 51012 160VA TRANSFORMATEUR TORIQUE 2X12V/6,66A
Yves.
Anyway . . .
Transformateurs - Transformateurs toriques - VELLEMAN 16012 51012 160VA TRANSFORMATEUR TORIQUE 2X12V/6,66A
Yves.
If your PSU is capable to give that current try 15R/10W in series with each tube
You need more current to heat the heater and a small voltage drop on res
After 20-30sec bypass
You need more current to heat the heater and a small voltage drop on res
After 20-30sec bypass
It would be far easier to use a bargained 24V 7A transformer with only a slight change to your filaments wiring. A parallel/series combination will do the trick, without a huge wires section. I've been using these satisfying combination for years.
My concern is space, there is no more space to put a transformer. I'll have to find a way to make this work. Besides, you can't beat the price. Cost me less than USD12 for the PSU. I'm sure it could work.. I just haven't found the way yet.
It seems my problem is not the overcurrent protection (as my PSU turns out has no such feature). It's actually the undervoltage protection. With 2 tubes or more, the voltage would drop and trigger the protection. My PSU uses EST7502 PWM chip and i have tried disabling the over/undervoltage protection and i got 24v at the 12v output (i think it's now an open-loop PSU as in there's no regulation done by the PWM IC). Tried connecting 4 tubes and it dropped to 18V. It's now a question of how to get 12V without tripping the undervoltage protection. Any ideas? I will try posting in other forums, especially SMPS forums.
It seems my problem is not the overcurrent protection (as my PSU turns out has no such feature). It's actually the undervoltage protection. With 2 tubes or more, the voltage would drop and trigger the protection. My PSU uses EST7502 PWM chip and i have tried disabling the over/undervoltage protection and i got 24v at the 12v output (i think it's now an open-loop PSU as in there's no regulation done by the PWM IC). Tried connecting 4 tubes and it dropped to 18V. It's now a question of how to get 12V without tripping the undervoltage protection. Any ideas? I will try posting in other forums, especially SMPS forums.
Could try an inrush current limiter, something like a CL-101, it has 0.5 ohms cold and 0.029 at 13.2 amps. Put it in series with all the rest of the filaments and the power supply will see 0.5 ohms plus the filaments. After it's warmed up the CL-101 will still be dropping 0.3 volts, unless you bypass it with a relay.
Last edited:
Toroid transformers are not more cumbersome than your SMPS.
My 300 VA is 110 mm diameter by 65 mm height...😉
My 300 VA is 110 mm diameter by 65 mm height...😉
SMPS is cheap and works great for heater supply. Just try it. I had with my 12v 13A heater(GU81m) only 25mV rimple at 50khz. Not measureble on the audio output.
You could start with a 24V filaments chain, heated under only 12V, and then switch back to 12V after about a minute, using a relay or even a manual switch.
With a simple monostable control, this could be an easy way to solve for your problem.
With a simple monostable control, this could be an easy way to solve for your problem.
I got from China a 350W 7,5 volts and I uses it with 20 6AS7G in parallel, they have a regulator that can change in 10% the voltage, I measured the voltage in each valve and is the same. The SMPS is not stressed at all. Run cold. 25 u$ give a big solution. There are for same power in 12, 24, 36 and 48 volts. Originally I thank in use a ATX smps but is a dirty cheap/expensive solution. They need to be charged if not no start, the chinnese smps has regulator, and short circuit protection and doesn't need to be charged for run.
It worked! Notice the past tense.. 😱
My suspicion was correct, the PSU failed to start because the undervoltage protection was working. The PWM IC on the PSU is EST7502 so i tried disabling the over/undervoltage protection by disconnecting pin 1,2,3 and 4 from the circuit (cut the PCB trace) and grounding them by soldering all to pin 5. This apparently worked.. for 30 minutes. After that, magic smoke decided to show up, leaving the PSU dead. There goes 12 bucks.
So, yes.. as mentioned by some earlier, my cheapo PSU doesn't seem to have the rating written on the box it was in. The EE35 core was very hot when the PSU died. I can't tell visually which component died but my guess is primary short in the EE35, it was very hot when the PSU was on. I read that EE35 can withstand 225VA though.. My load was around 150VA, so what happened?
Next, higher wattage PSU or the good-ol' transformer?
An externally hosted image should be here but it was not working when we last tested it.
My suspicion was correct, the PSU failed to start because the undervoltage protection was working. The PWM IC on the PSU is EST7502 so i tried disabling the over/undervoltage protection by disconnecting pin 1,2,3 and 4 from the circuit (cut the PCB trace) and grounding them by soldering all to pin 5. This apparently worked.. for 30 minutes. After that, magic smoke decided to show up, leaving the PSU dead. There goes 12 bucks.
So, yes.. as mentioned by some earlier, my cheapo PSU doesn't seem to have the rating written on the box it was in. The EE35 core was very hot when the PSU died. I can't tell visually which component died but my guess is primary short in the EE35, it was very hot when the PSU was on. I read that EE35 can withstand 225VA though.. My load was around 150VA, so what happened?
Next, higher wattage PSU or the good-ol' transformer?
Can you provide link to the seller?
Now there are zillons of them on ebay, this was three year ago, anyway I will try, IE I attach something I looked for by ebay: Transformator DC 12V 29A 350W Netzteil Schaltnetzteil Stromversorgung Adapter | eBay
I won't give up on SMPS just yet. It's weekend and i'm going to see if i can revive that dead ATX PSU. I'm going to put better ventilation this time by installing the fan somewhere on my chassis. If that doesn't work, i already have a custom SMPS diptrace file waiting to be sent to the PCB fabricator and an ETD39 core waiting to be wound. It would be my first time making SMPS so if this works i'd be very happy.
Oye amigo, The SMPS I described bring a fan, but if not uses it, Doesn't work hot at all.
I know is romantic the ATX because they are free. But I don't imagine how is the result of this.
Anyway I hope the best for you friend
I know is romantic the ATX because they are free. But I don't imagine how is the result of this.
Anyway I hope the best for you friend
4 tubes. you must have (3.6Ax4)=14.4A 14.4*100=1440 1440/80=18A you must have a 12V rail that is 18A or better or it will crowbar on power up. Also if you are using an atx, use the green wire on the atx connector grounded as the power switch. turning the supply on/off with the vacation switch and the green wire grounded can sometimes trigger the protection circuit
yes...from the picture posted, the atx used was a generic one and overrated, two or more of those would have been better and get an atx that is certified 80%, 80 Plus - Wikipedia, the free encyclopedia
two of those atx's for 24volts and heaters wired for 24 volts, (series - parallel) will be easier to load...
- Status
- Not open for further replies.