I'm playing with some Nixie tubes and needed a cheap high voltage supply from DC so purchased a couple of the ebay "DC-DC 8~32V to 45~390V High Voltage Boost Converters". These work fine with the Nixie so I thought I would have a look at whether they could be used with simple SPUD type amps.
To test I set the output at 285VDC and used a 25Watt 5K6 resistor as the load modelling a single channel SE amp. The set voltage regulation was not affected by connecting the resistor - so far so good.
The values chosen give an output current of 50.9mA and power of 14.5 Watts (well within the advertised 40W average 70W peak rating). Input power was 13.8V at 1.23A = 17W so an efficiency of 85% - not too bad. With a run time of 10 minutes nothing on the convertor got remotely hot (unlike the resistor!). The switching controller is a UC3843A.
Switching frequency was a rock solid 100KHz. As can be seen the switching is creating a spike of just over a volt on the output but this is with an output capacitor of just 10uF on the PCB and no filtering.
To test I set the output at 285VDC and used a 25Watt 5K6 resistor as the load modelling a single channel SE amp. The set voltage regulation was not affected by connecting the resistor - so far so good.
The values chosen give an output current of 50.9mA and power of 14.5 Watts (well within the advertised 40W average 70W peak rating). Input power was 13.8V at 1.23A = 17W so an efficiency of 85% - not too bad. With a run time of 10 minutes nothing on the convertor got remotely hot (unlike the resistor!). The switching controller is a UC3843A.
Switching frequency was a rock solid 100KHz. As can be seen the switching is creating a spike of just over a volt on the output but this is with an output capacitor of just 10uF on the PCB and no filtering.
Attachments
Funny, it's advertised as being 75kHz.
In my experience it's only good for about 10W in an enclosed chassis (minimal cooling). almost 20W if you change the diodes for UF4007.
I'd use either of these two for a SPUD.
MINI DC-AC Inverter 12V to 18V220V/380V 500W Boost Step UP Power Module New Hot | eBay
DC-AC Converter 12V to 110V 200V 220V 280V 150W Inverter Boost Board Transformer | eBay
The bonus is they are both PP, not SE. I have a 20WPC integrated amp with 22 tubes running off these.
The one you listed is good for a line stage, buffer, or screen supply though.
In my experience it's only good for about 10W in an enclosed chassis (minimal cooling). almost 20W if you change the diodes for UF4007.
I'd use either of these two for a SPUD.
MINI DC-AC Inverter 12V to 18V220V/380V 500W Boost Step UP Power Module New Hot | eBay
DC-AC Converter 12V to 110V 200V 220V 280V 150W Inverter Boost Board Transformer | eBay
The bonus is they are both PP, not SE. I have a 20WPC integrated amp with 22 tubes running off these.
The one you listed is good for a line stage, buffer, or screen supply though.
The diode on my boards are US3M which are at least as well specced as the UF4007. I imagine you get a different component selection depending on what was on hand when they were fabricated - pot-luck in other words
I've had them overheat and blow up (shorted, but the board doesn't seem to care). The UF4007 is physically bigger, especially if you don't trim the leads and in theory dissipates heat better. I couldn't even find data on the US3M... The thermal resistance of the UF4007 is slightly better, too.
Keep in mind I run them in ~50c ambient.
Keep in mind I run them in ~50c ambient.
kodamax, you had poor performance from a similar module due to choke filtering the output:
'150W' 12VDC isolated 25kHz step-up smps for B+
Did you get poor performance with the module in this link for the same reason?
I'm surprised you are getting this module to blow up! I'm with matherp on the nice efficiency and low temperature rise performance.
'150W' 12VDC isolated 25kHz step-up smps for B+
Did you get poor performance with the module in this link for the same reason?
I'm surprised you are getting this module to blow up! I'm with matherp on the nice efficiency and low temperature rise performance.
I was also dyslexic. I was looking for USM3. I no longer use a choke input.
Put this module in a metal box with no (or minimal) vents and ask it for 330V, 50ma. It will last about 30 minutes. The diodes go first, if you replace them with better ones either the FET or the transformer will burn out unless you use forced air cooling. The last one that blew was about 90c by the time I had the cover off of the phono amp.
Put this module in a metal box with no (or minimal) vents and ask it for 330V, 50ma. It will last about 30 minutes. The diodes go first, if you replace them with better ones either the FET or the transformer will burn out unless you use forced air cooling. The last one that blew was about 90c by the time I had the cover off of the phono amp.
Any such equipment will overheat if you significantly restrict cooling ability. I'd suggest that is not the module's 'fault', but your poor awareness of how to transfer heat away from the module when put in that enclosure.
I call it derating... This module would pass 100W if it was cooled by LN2, that's not the point. In a typical tube amp chassis the ambient temperature is easily 50c. In an audio application you don't want a cooling fan adding noise, either.
Plus, a 40C rise for a 15W load is attrocious.
Plus, a 40C rise for a 15W load is attrocious.
The ps module has a heatsink - if you use the module you have to manage cooling that heatsink - if you can't manage that activity then its a poor practical implementation, not the fault of the module. Telling everyone that the module is crap is misleading, as it's your practical implementation that is failing.
Stick any semi device on a heatsink, and then restrict natural airflow around the heatsink by enclosing it in a box indicates poor awareness of thermal management of semiconductor devices.
If enclosing in a box is the requirement then you need to modify the thermal transfer path and imho add a conductive transfer to the box and ensure the box can conduct heat to the chassis or where external ambient air can provide sufficient natural circulation to transfer the heat away. Your box is presently limiting convective cooling capability of the heatsink.
There are many good tutorials on thermal design and management of semiconductor devices on-line.
Stick any semi device on a heatsink, and then restrict natural airflow around the heatsink by enclosing it in a box indicates poor awareness of thermal management of semiconductor devices.
If enclosing in a box is the requirement then you need to modify the thermal transfer path and imho add a conductive transfer to the box and ensure the box can conduct heat to the chassis or where external ambient air can provide sufficient natural circulation to transfer the heat away. Your box is presently limiting convective cooling capability of the heatsink.
There are many good tutorials on thermal design and management of semiconductor devices on-line.
Why would I re-invent the wheel when I can just use one of the other modules I've refered to?
As mentioned the failure mode isn't primarily the FET. It's the diodes followed by the windings of the transformer. Indeed the heatsink is ~20c cooler than the transformer. I'm not trying to mislead anyone, and I didn't refer to this module anywhere as crap. I'm simply stating my observations in a practical implementation in the average environment you would find in a tube amp chassis using only convective cooling. I built a preamp with one of these that only uses 300V/20mA and it's been solid for 3 years, I built a phono stage that needed 335V/50ma and it couldn't handle it. Much like the Chinese 720W SMPS that says 500W on the transformer inside, these devices are rated optimistically.
As mentioned the failure mode isn't primarily the FET. It's the diodes followed by the windings of the transformer. Indeed the heatsink is ~20c cooler than the transformer. I'm not trying to mislead anyone, and I didn't refer to this module anywhere as crap. I'm simply stating my observations in a practical implementation in the average environment you would find in a tube amp chassis using only convective cooling. I built a preamp with one of these that only uses 300V/20mA and it's been solid for 3 years, I built a phono stage that needed 335V/50ma and it couldn't handle it. Much like the Chinese 720W SMPS that says 500W on the transformer inside, these devices are rated optimistically.
Last edited:
If you swapped out the US3M surface mount diode for a leaded UF4007 then that makes total sense for a more failure prone module !
The UF4007 has a much lower current rating; 1A versus 3A average, and the 3A is at 75C track compared to the 1A at 55C ambient with standard lead length clamping. And the surface mount heat removal should be better than the leaded package. And the US3M forward voltage spec sure looks less, with 1.2V at 1A at 25C, versus 1.5V. For a flyback sawtooth current waveform, that all mounts up to way more diode stress on a UF4007.
The UF4007 has a much lower current rating; 1A versus 3A average, and the 3A is at 75C track compared to the 1A at 55C ambient with standard lead length clamping. And the surface mount heat removal should be better than the leaded package. And the US3M forward voltage spec sure looks less, with 1.2V at 1A at 25C, versus 1.5V. For a flyback sawtooth current waveform, that all mounts up to way more diode stress on a UF4007.
- Status
- Not open for further replies.