I've had good success using these inexpensive DC-DC converters for vacuum tube B+ for a couple of years, but recently had two failures. A previous thread here reverse-engineering the circuit was very helpful but I wonder if a more complete schematic is available, as a couple of links are no longer working.
I'm going to do more measurements but have one of the units working again after replacing the FET. The RU7088 used is a 65 volt device and I'm going to see how close to that limit the drain voltage gets with the supply cranked up to full voltage, as I was had done. Another FET tested good using a DMM but still would not work in the board, although an IRF510 replacement does.
I also found the half-wave rectifier diode was leaky causing a low impedance at the output, and am puzzled about how that happened and whether it could have loaded the secondary enough to cause the FET to fail. I would have expected the overcurrent protection to kick in but some boards have a metal shorting bar in place of the current sense resistor.
I just thought I'd ask to see if anyone has experienced similar failures with these inexpensive boards. I'm also interested in any open sourced 3-400 volt DC-DC or line operated switchers good for 100mA for vacuum tube projects if anyone has suggestions (ideally using off-the-shelf magnetics).
Thanks, Bob
I'm going to do more measurements but have one of the units working again after replacing the FET. The RU7088 used is a 65 volt device and I'm going to see how close to that limit the drain voltage gets with the supply cranked up to full voltage, as I was had done. Another FET tested good using a DMM but still would not work in the board, although an IRF510 replacement does.
I also found the half-wave rectifier diode was leaky causing a low impedance at the output, and am puzzled about how that happened and whether it could have loaded the secondary enough to cause the FET to fail. I would have expected the overcurrent protection to kick in but some boards have a metal shorting bar in place of the current sense resistor.
I just thought I'd ask to see if anyone has experienced similar failures with these inexpensive boards. I'm also interested in any open sourced 3-400 volt DC-DC or line operated switchers good for 100mA for vacuum tube projects if anyone has suggestions (ideally using off-the-shelf magnetics).
Thanks, Bob
I've heard of a few that have had these die. Part of the reason for me trying to DIY my own is (a) I know what goes in it - parts and design (b) it's open and (c) the parts are OTS such as Coilcraft.
I looked and they all seem to follow the same image but don't have any real explanation of features or power abilities vs capacitance vs noise.
Only issue is as soon as you get to OTS, you're looking at more expense. I've deliberately gone down the route of a OTS trusted 240ac -> 24Vdc with isolation, safety, filtering, PFC etc, then make something simple that goes from 24Vdc to what I want with minimal parts (ignoring the capacitors).
I looked and they all seem to follow the same image but don't have any real explanation of features or power abilities vs capacitance vs noise.
Only issue is as soon as you get to OTS, you're looking at more expense. I've deliberately gone down the route of a OTS trusted 240ac -> 24Vdc with isolation, safety, filtering, PFC etc, then make something simple that goes from 24Vdc to what I want with minimal parts (ignoring the capacitors).
Last edited:
https://www.dalmura.com.au/static/YH11068A.pdf
https://forum.cxem.net/uploads/monthly_2020_06/yh11068a_boost_converter-v1.png.67b1e72e6d9f03963ab9ecbb91980004.png
I'd suggest your failures relate to you inappropriately operating the smps - by that I mean you may have not managed the thermal dissipation requirements (especially for the FET), or any sustained high output current, or operated at too high an output voltage (which could stress the output capacitor and lead to stressed diode), or used at too low or high an input voltage. Although the input and output voltage ranges are identified, unless you perform your own reverse engineering assessment of operation, I'd suggest you provide significant margin. There are many ways the smps can be stressed and fail, not just Vds.
There have been quite a few other smps pcb modules suitable for diy valve projects. The better ones I have found start out as 12V nominal inverters (with ebay titles like "150W Inverter Boost Board Transformer DC-AC Converter 12V to 110V 200V 220V 280V") for AC output powering, and the output circuit needs to be modified to provide a suitable filtered B+ - they are typically unregulated but if the input DC is regulated/stiff, then the B+ level is similarly stiff.
'150W' 12VDC isolated 25kHz step-up smps for B+
There may also still be 12VDC to AC car adaptors, that can be fairly easily converted to a fixed B+ output by modifying the transformer secondary side circuitry.
75W car inverter.pdf
https://forum.cxem.net/uploads/monthly_2020_06/yh11068a_boost_converter-v1.png.67b1e72e6d9f03963ab9ecbb91980004.png
I'd suggest your failures relate to you inappropriately operating the smps - by that I mean you may have not managed the thermal dissipation requirements (especially for the FET), or any sustained high output current, or operated at too high an output voltage (which could stress the output capacitor and lead to stressed diode), or used at too low or high an input voltage. Although the input and output voltage ranges are identified, unless you perform your own reverse engineering assessment of operation, I'd suggest you provide significant margin. There are many ways the smps can be stressed and fail, not just Vds.
There have been quite a few other smps pcb modules suitable for diy valve projects. The better ones I have found start out as 12V nominal inverters (with ebay titles like "150W Inverter Boost Board Transformer DC-AC Converter 12V to 110V 200V 220V 280V") for AC output powering, and the output circuit needs to be modified to provide a suitable filtered B+ - they are typically unregulated but if the input DC is regulated/stiff, then the B+ level is similarly stiff.
'150W' 12VDC isolated 25kHz step-up smps for B+
There may also still be 12VDC to AC car adaptors, that can be fairly easily converted to a fixed B+ output by modifying the transformer secondary side circuitry.
75W car inverter.pdf
Thanks for the comments. I agree that I did something to kill these two because I've taken no special precautions in the past and they've held up fine. So I'll be double-checking everything. Good suggestion on the inverter modules, I've been buying and trying various types but had kinda converged on these.
For the benefit of anyone coming across this thread in future I'll add that both half-wave diodes were shorted or badly leaky, which from what I've read usually is caused by a large reverse spike, so that's a good clue. I also will admit to having "cranked it up to 11", i.e. I set the pot for maximum voltage, and the filter cap is just rated at 400 VDC. I'm guessing, but it could have been a cascade failure where the output short caused when the diode shorted put enough load on the primary that the FET stopped switching, or the control signal went high which biased the FET on long enough for it to draw excess current. In both cases the 10A minifuse was blown and the FET is the only thing capable of drawing enough current to do that (my power supply can source 20A and has short circuit protection, but evidently the fuse was quicker).
Bob
For the benefit of anyone coming across this thread in future I'll add that both half-wave diodes were shorted or badly leaky, which from what I've read usually is caused by a large reverse spike, so that's a good clue. I also will admit to having "cranked it up to 11", i.e. I set the pot for maximum voltage, and the filter cap is just rated at 400 VDC. I'm guessing, but it could have been a cascade failure where the output short caused when the diode shorted put enough load on the primary that the FET stopped switching, or the control signal went high which biased the FET on long enough for it to draw excess current. In both cases the 10A minifuse was blown and the FET is the only thing capable of drawing enough current to do that (my power supply can source 20A and has short circuit protection, but evidently the fuse was quicker).
Bob
I wanted to close the loop on this issue, starting by saying thanks for the comments and suggestions, and especially for the link to the updated schematic and to whoever took the time to reverse-engineer the board. Like most here I suspect, I'm not very good at black-boxes, I want to know how it works, and the schematic was very helpful.
I've also confirmed the problem was self-inflicted, as I figured it would be since I've used these modules for some time but these failures occurred in a new power supply assembly. The reason for the failure isn't very obvious since an ohmmeter check will confirm that the DC input "-" and HV output "-" terminals are connected together, and the schematic shows this to be true also. So, for no good reason I decided to connect a single ground wire from my 12 volt supply to the HV "-" terminal. After all, "ground is ground", right?
Well yes, but we also know that current flow through PCB traces is not like a piece of #12 copper wire, and evidently by forcing the DC current from the power supply to flow through the traces associated with the output circuit, the UC3843 would stop oscillating, or maybe it just upset the control voltage inputs, but the effect was to cause the FET to short and blow the fuse. I'm not sure why but the HV rectifier also shorted, and I've also decided to reduce the output voltage to no more than 350VDC to ensure a safety margin. My new supply is back in operation after wiring the board up the way it's supposed to be, so while I'm not 100% sure of the root cause, the bottom line is the failures don't seem to be due to any defect in the board.
Bob
I've also confirmed the problem was self-inflicted, as I figured it would be since I've used these modules for some time but these failures occurred in a new power supply assembly. The reason for the failure isn't very obvious since an ohmmeter check will confirm that the DC input "-" and HV output "-" terminals are connected together, and the schematic shows this to be true also. So, for no good reason I decided to connect a single ground wire from my 12 volt supply to the HV "-" terminal. After all, "ground is ground", right?
Well yes, but we also know that current flow through PCB traces is not like a piece of #12 copper wire, and evidently by forcing the DC current from the power supply to flow through the traces associated with the output circuit, the UC3843 would stop oscillating, or maybe it just upset the control voltage inputs, but the effect was to cause the FET to short and blow the fuse. I'm not sure why but the HV rectifier also shorted, and I've also decided to reduce the output voltage to no more than 350VDC to ensure a safety margin. My new supply is back in operation after wiring the board up the way it's supposed to be, so while I'm not 100% sure of the root cause, the bottom line is the failures don't seem to be due to any defect in the board.
Bob
I have fried two of those and have the impression that the voltage feedback of the booster sometimes doesn't work. The voltage hits nearly 600V and destroys the 400V capacitor in seconds..
I had planned to use those in my 3-channel high gain tube preamp and did use a common ground and not the negative (ground) output of the board. There is a few ohms between the ground input and the output of the board, which possibly causes a nasty ground loop. I have used and tested them with the common ground for days. And they worked fine until I have stuck them in an aluminium enclosure.. I think the failures might have something to do with the effect of the enclosure on the transformer..
I still (try to) use one of those in my preamp.. If it starts working allright, the voltage stays at the 350V without any problems.. If it doesn't start right, I switch off the amp and switch it on again.. I just watch the current, when I am switching on.. If the current is too high, I turn it immediately off. The board starts right %30-40 of the time..
Using 555 and UC3843 I tried to build my own voltage boosters using the transformer from one of the failed boards.. No success till now.. My voltages are under 300V and get much lower with a load.. It possibly won’t be easy.
I had planned to use those in my 3-channel high gain tube preamp and did use a common ground and not the negative (ground) output of the board. There is a few ohms between the ground input and the output of the board, which possibly causes a nasty ground loop. I have used and tested them with the common ground for days. And they worked fine until I have stuck them in an aluminium enclosure.. I think the failures might have something to do with the effect of the enclosure on the transformer..
I still (try to) use one of those in my preamp.. If it starts working allright, the voltage stays at the 350V without any problems.. If it doesn't start right, I switch off the amp and switch it on again.. I just watch the current, when I am switching on.. If the current is too high, I turn it immediately off. The board starts right %30-40 of the time..
Using 555 and UC3843 I tried to build my own voltage boosters using the transformer from one of the failed boards.. No success till now.. My voltages are under 300V and get much lower with a load.. It possibly won’t be easy.
I usee these instead now.
https://www.aliexpress.com/item/1005001756171525.html?spm=2114.13010708.0.0.29ab4c4dj91k66If you need more than 280V, use the AC output and use a Delon doubler. If you do this, remember to use fast diodes (like HER208).
https://www.aliexpress.com/item/1005001756171525.html?spm=2114.13010708.0.0.29ab4c4dj91k66If you need more than 280V, use the AC output and use a Delon doubler. If you do this, remember to use fast diodes (like HER208).