For max voltage limiting I settled on two cheap GDT tubes in parallel with the halves of the primary OPT winding. The GDT will be in series with a current quenching resistor.
2095-350-BLF - BOURNS - GAS DISCHARGE TUBE, 2P, 3500V, 20% | Newark element14
Voltage breakdown is 3.5kv 20%.
One of the issues at the moment is calculating the the value of the resistor. Thinking of experimenting with 360K 50W.
Please feel free to give me your suggestions. That would've been my next question
2095-350-BLF - BOURNS - GAS DISCHARGE TUBE, 2P, 3500V, 20% | Newark element14
Voltage breakdown is 3.5kv 20%.
One of the issues at the moment is calculating the the value of the resistor. Thinking of experimenting with 360K 50W.
Please feel free to give me your suggestions. That would've been my next question
I'm a little late to the party... but that's what I designed this thing for:
Isolated Current Sensor
I used it to sense the plates directly on an 815 tube.
All it takes is the current sensor and one opamp per tube.
Pete
Isolated Current Sensor
I used it to sense the plates directly on an 815 tube.
All it takes is the current sensor and one opamp per tube.
Pete
What is your max design OT CT DC voltage ? Discharge tubes are a bit abrupt, and I guess may continue to conduct when high frequency AC is applied.
What is your likely V-I characteristic of the power supply if the primary half-windings are effectively shorted ?
You may also want to consider a softer clipping mechanism prior to the discharge tubes kicking in - such as a MOV string - that may minimise when the discharge tubes are needed.
What is your likely V-I characteristic of the power supply if the primary half-windings are effectively shorted ?
You may also want to consider a softer clipping mechanism prior to the discharge tubes kicking in - such as a MOV string - that may minimise when the discharge tubes are needed.
Oh yes, I remember reading that article well! Impressive work with the panel. As I recall, however, no indication of the sensor's max voltage was available.
My working voltage, with all tubes drawing their idle current will be of around 1600v
The Transformer itself is capable of almost 2A of current on its anode line (huge semi-industrial surplus).
I thought about MOVs, not a bad idea. Either with GDTs or MOVs I would still need some kind of limiting resistor, correct?
with GDT the resistor would act as an arc breaker. With MOVs I think it could help prevent a small explosion.
My working voltage, with all tubes drawing their idle current will be of around 1600v
The Transformer itself is capable of almost 2A of current on its anode line (huge semi-industrial surplus).
I thought about MOVs, not a bad idea. Either with GDTs or MOVs I would still need some kind of limiting resistor, correct?
with GDT the resistor would act as an arc breaker. With MOVs I think it could help prevent a small explosion.
Yeah, that bothered me as well. I think somehow I came to the conclusion it was good for 1500V, but I'm not sure. At 1600V, you would probably be safer with a more expensive sensor.
As was mentioned there are quite a few to choose from, like those from LEM.
Pete
Yes, that should work. I guess you still need to consider the bobbin material, but I suspect it would be fine...
Not in my view - sensing in cathode side is also practical, and safer (especially for this application).
That's my view as well, particularly given the high B+ voltage... Since the tubes are AC heated, wouldn't there be a need for some hum-bucking adjustment on the cathodes as well? If so, then it makes sense to implement low-side sensing. But the slew of high-side sensing proposals above seem to suggest otherwise, so just wondering if I missed something...
I'd suggest it was a mistake.
I like the idea of the inherent neutralisation of heater hum with a quad. I think a few of us just assumed you would get around to using a humdinger, so we presumed certain sensor/meter techniques to address that - it would have assisted some posters to know that up front, as your schematic doesn't indicate that.
Perhaps it would also be good to know what metering you had in mind, as that can also twist and turn the assistance you are posting for.
My schematic deliberately lacks indication of a humdinger because I am not using any. I suppose I could've stressed this more but I wasn't expecting that an assumption would be made that a humdinger was installed.
I suppose a humdinger could work provided that it had adequate resistors to reduce total current.
I was very interested in your suggestion for opt protection. Would you like to share more info? Maybe in pvt since it is likely off topic.
As I said my initial thoughts were to sense cathode current to allow equalization of tube idle current and avoid as far as possible transformer saturation. Whether I sense current or a voltage it is non consequential as long as I can get close to reducing disparity between tube pairs.
My idea was to either sense voltage drop across a 1ohm resistor in the anode or cathode or somehow sense current directly at either of those two points. I do like the simplicity of just splitting the transformer and get current measurement with a run-of-the-mill panel current meter.
Inviato dal mio GT-I9301I utilizzando Tapatalk
I suppose a humdinger could work provided that it had adequate resistors to reduce total current.
I was very interested in your suggestion for opt protection. Would you like to share more info? Maybe in pvt since it is likely off topic.
As I said my initial thoughts were to sense cathode current to allow equalization of tube idle current and avoid as far as possible transformer saturation. Whether I sense current or a voltage it is non consequential as long as I can get close to reducing disparity between tube pairs.
My idea was to either sense voltage drop across a 1ohm resistor in the anode or cathode or somehow sense current directly at either of those two points. I do like the simplicity of just splitting the transformer and get current measurement with a run-of-the-mill panel current meter.
Inviato dal mio GT-I9301I utilizzando Tapatalk
It may be heretical to say so, but you can use ordinary d'Arsonval meters in the anode lines if they're safely boxed in plexi, or similar. I recently went through this same decision tree for a triple 211 parallel amp; liked optocouplers but ended up with separate filament transformers (had 'em on-hand anyway). Less grief all around.
All good fortune,
Chris
All good fortune,
Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.