Just in case, here is a suitable kit for speaker protection: https://www.aliexpress.us/item/3256802693700015.html
The cap on pin 7 on the UPC 1237 needs to be increased to enable longer delay. 1000uF should do fine. But you may have to experiement. Just make the delay long enough for the tubes to come up to temp. Around 30 seconds to 60 seconds should work.
What is nice about the UPC 1237 is that it will trigger the relay if too much DC is seen on the speaker terminals, - thus protecting your speakers.
The cap on pin 7 on the UPC 1237 needs to be increased to enable longer delay. 1000uF should do fine. But you may have to experiement. Just make the delay long enough for the tubes to come up to temp. Around 30 seconds to 60 seconds should work.
What is nice about the UPC 1237 is that it will trigger the relay if too much DC is seen on the speaker terminals, - thus protecting your speakers.
Seems like the thread is drifting further away from Circlotron topic.
Regarding speaker protection, none of Circlotron outputs is ground therefore the one mentioned above might not work the best IMHO. You can use each stereo board to monitor both outputs of one channel, but that's respect to Ground not to each other.
Regarding speaker protection, none of Circlotron outputs is ground therefore the one mentioned above might not work the best IMHO. You can use each stereo board to monitor both outputs of one channel, but that's respect to Ground not to each other.
Speaker protection does indeed work for circlotron. You just need two of them. Remember that the Circlotron has a ground bwteen the the two speaker terminals.
Here is the protection circuit added:
In fact, I used this very arrangement in the amplifier I brought to Burning Amp 2023. One relay per channel (instead of one relay for 2 channels).
If more than a volt or so is seen across either 50 ohm resistor, the pretection will trigger. But more importantly, with a large cap on pin 7 of the UPC 1237, you get a nice delay on power on.
The positive phase and and the negative phase can be driven by two 6DJ8s:
In fact, I used this very arrangement in the amplifier I brought to Burning Amp 2023. One relay per channel (instead of one relay for 2 channels).
If more than a volt or so is seen across either 50 ohm resistor, the pretection will trigger. But more importantly, with a large cap on pin 7 of the UPC 1237, you get a nice delay on power on.
The positive phase and and the negative phase can be driven by two 6DJ8s:
Yes like I mentioned it does work (and I did use it before) however one need to remember the 1237 monitor each output to Ground, not between the Outputs.
And it does. When you use a stereo protection circuit where L and R is used for + and - of the Circlotron output. In fact, the Circlotron can be viewed as a half-wave bridge amplifier. Chassis ground is formed by the two 50 ohm resistors between speaker + and speaker -.
So, you are right: in the Circlotron the protection circuit behaves as it would with a stereo amp in bridged mode. Ground is floating between + and - .
So, you are right: in the Circlotron the protection circuit behaves as it would with a stereo amp in bridged mode. Ground is floating between + and - .
OK fellow amplifier builders, - and since this is a Circlotron thread - here is Blue Smoke 11. A circlotron MOSFET with a tube driver. Please note, this amplifier needs a balanced input signal. 😎 :
Please note that you need 5 power supplies here. 2 floating ones for the Circlotron output and one B+ for the tubes. Then a heater supply for the tubes - which needs to be elevated at +60V or so becasue we are running SRPP here. And finally seperate one for the BIAS supply for the MOSFETs.
Please note that you need 5 power supplies here. 2 floating ones for the Circlotron output and one B+ for the tubes. Then a heater supply for the tubes - which needs to be elevated at +60V or so becasue we are running SRPP here. And finally seperate one for the BIAS supply for the MOSFETs.
Denis
Thanks for the Blue Smoke 11, looks very inviting. So much so, that I have ordered the DC protection and SRPP boards you suggested. Was wondering about adjusting for DC offset, was thinking individual bias pots would be a solution. Or, were you thinking of something else?
Thanks for the Blue Smoke 11, looks very inviting. So much so, that I have ordered the DC protection and SRPP boards you suggested. Was wondering about adjusting for DC offset, was thinking individual bias pots would be a solution. Or, were you thinking of something else?
Hi Ejam,
Sure, we can add DC offset. It may not be needed - depending on how well the MOSFETs are matched. But here you go:
Sure, we can add DC offset. It may not be needed - depending on how well the MOSFETs are matched. But here you go:
Ratbaggp
Bluesmoke 11 is a concept to date, although Denis has constructed similar but higher powered circlotrons (his Burning Amp 2023 Circlotron). Something like a Hafler DH200 heatsinks would be fine. The chassis's dimensions are 5 1/8 x 16 x 10 1/2 inches (HxWxD), to give you an idea of the heatsink size, photo included. If building from scratch, Conrad Heatsinks model MF30-1F-75 at 0.37 degrees/W would be more than adequate.
Bluesmoke 11 is a concept to date, although Denis has constructed similar but higher powered circlotrons (his Burning Amp 2023 Circlotron). Something like a Hafler DH200 heatsinks would be fine. The chassis's dimensions are 5 1/8 x 16 x 10 1/2 inches (HxWxD), to give you an idea of the heatsink size, photo included. If building from scratch, Conrad Heatsinks model MF30-1F-75 at 0.37 degrees/W would be more than adequate.
Attachments
Wow - you beat me to it 😎
Heatsink is similar to HAFLER 200/220/Pro2400 etc. - or even B&K ST140 in a pinch.
Had its bigger brother at BAF 2023 with 6 MOSFETs per channel.
Here is a box that would work: https://www.aliexpress.us/item/3256802402964067.html
I woud at least use 4 MOSFETs per channel.
Also had a non-circlotron at Burning Amp 2024:
Heatsink is similar to HAFLER 200/220/Pro2400 etc. - or even B&K ST140 in a pinch.
Had its bigger brother at BAF 2023 with 6 MOSFETs per channel.
Here is a box that would work: https://www.aliexpress.us/item/3256802402964067.html
I woud at least use 4 MOSFETs per channel.
Also had a non-circlotron at Burning Amp 2024:
The protection time delay needs to be increased. The cap on pin 7 must be increased. Shoot for ~20 seconds in real life. The actual value depends on the resistor value feeding that pin. See here: https://ampslab.com/PDF/upc1237.pdf
I like ~ 470uF here. Plenty of delay for the tubes to warm up when the amp is first turned on. And it does not hurt of someting trips later.
I like ~ 470uF here. Plenty of delay for the tubes to warm up when the amp is first turned on. And it does not hurt of someting trips later.
Aliexpress or Ebay have many examples of little kits with the UPC 1237. Here is a typical schematic:
The 56K sensor resistors on pin two do the detecting of any large DC offset on output terminals of the amp. For circlotron we use both L and R for one channel - as the output is floating.
Pin 7 sets the delay. So here 330uF or so will 10X the delay. You can use anything here from 330uF to 1000 uF or larger. The bigger the cap, the longer the delay.
The 56K sensor resistors on pin two do the detecting of any large DC offset on output terminals of the amp. For circlotron we use both L and R for one channel - as the output is floating.
Pin 7 sets the delay. So here 330uF or so will 10X the delay. You can use anything here from 330uF to 1000 uF or larger. The bigger the cap, the longer the delay.
Yes, agreed. But a 2 second delay on the speaker protection - in case of DC on the speaker leads, seems too long.
Hello Vilfort,
I've never seen or heard you here before. Anyway very interested in your Hybrid power amp. Is there a thread dedicated to it?? I have always been looking at a tube front end for an amp. So the schematics are finished, all components shown except for power supply? thanks for presenting this.
I've never seen or heard you here before. Anyway very interested in your Hybrid power amp. Is there a thread dedicated to it?? I have always been looking at a tube front end for an amp. So the schematics are finished, all components shown except for power supply? thanks for presenting this.
About the delay: we are talking power on delay or post triggered delay. Not the trigger delay. The sensor pin 2 is what triggers. Then post trigger event, the time constant on pin 7 is what sets the delay before the speakers are turned back on. You can test this on the bench with a UPC 1237 for yourself. It triggers fast.
About a general discussion on hybrid amps, I gave a talk at Burning Amp 2019:
About a general discussion on hybrid amps, I gave a talk at Burning Amp 2019:
I am more interested in the protection circuit.
If a tube shorts, how long does it take to disconnect the speakers? I would hope for small fractions of a second.
Sorry for not being clear. I have a "slow start" on mine now. I turn on the main power wait for at least a minute and switch the B+ power supplies on. (Plate voltage + rail and - rail to the 16 output tubes). AND in about 2 seconds the B+ and (-) come up slowly - great and recommended.
What I was looking for was a quick speaker safety disconnect. Though mine are fused and that does work, for less than $10 it would be good insurance.
This is probably more important on a Futterman than a circlotron.
Russ
If a tube shorts, how long does it take to disconnect the speakers? I would hope for small fractions of a second.
Sorry for not being clear. I have a "slow start" on mine now. I turn on the main power wait for at least a minute and switch the B+ power supplies on. (Plate voltage + rail and - rail to the 16 output tubes). AND in about 2 seconds the B+ and (-) come up slowly - great and recommended.
What I was looking for was a quick speaker safety disconnect. Though mine are fused and that does work, for less than $10 it would be good insurance.
This is probably more important on a Futterman than a circlotron.
Russ