I guess guys like Dave Reich of Theta Digital beg to differ in this respect, at least with respect to the NC1200 modules.
Sami, Rens - I am very interested in your work and its applicability to the Hypex NC400 modules. Any immediate thoughts in this regard while I keep digesting your work thus far?
I know your focus thus far, and the core of this thread, is to discuss the circuit for DC Error-driven interruption of the main power rails, but I have some immediate questions on the rest of your schematic:
- the UcD1200 calls for +/- 75-95V power supply voltage (+/-90V typical), +/-12V Audio Input/Buffer power supply and +15V driver supply voltage. 55VAC for the power supply voltage AC would seem to low, no?
- are their significant benefits to be obtained by providing better regulation of the buffer supply voltage and the driver supply voltage? The current needs of these rails are relatively modest (e.g. max 200mA for the driver supply) and so implementation of better regulation than using an LM317 or is relatively easy (e.g. see this thread for an example of improving on an LM317)
- IC7 can be a LM317?
- diodes D28 and D29, I thought these were only necessary if there was a cap from LM317 ref to GND (which itself would improve regulation)
- C7, C8, C27 and C28 are caps across the secondaries but won't do much for transformer ringing. Not better to allow for a proper CRC snubber? (see here)
Sorry if some of these questions are stupid or "off target". I'm very interested in seeing this move forward.
- the UcD1200 calls for +/- 75-95V power supply voltage (+/-90V typical), +/-12V Audio Input/Buffer power supply and +15V driver supply voltage. 55VAC for the power supply voltage AC would seem to low, no?
- are their significant benefits to be obtained by providing better regulation of the buffer supply voltage and the driver supply voltage? The current needs of these rails are relatively modest (e.g. max 200mA for the driver supply) and so implementation of better regulation than using an LM317 or is relatively easy (e.g. see this thread for an example of improving on an LM317)
- IC7 can be a LM317?
- diodes D28 and D29, I thought these were only necessary if there was a cap from LM317 ref to GND (which itself would improve regulation)
- C7, C8, C27 and C28 are caps across the secondaries but won't do much for transformer ringing. Not better to allow for a proper CRC snubber? (see here)
Sorry if some of these questions are stupid or "off target". I'm very interested in seeing this move forward.
Hi, 55Vac would make about +/-78Vdc. I'have two separate 730VA trafos with 2x55Vac and only one 1kVA 2x60 (+/-85Vdc). I like to have dual mono, this is why I decided to choose 55Vac voltage. Although, for UCD400 I have 2 trafos with 2x38Vac which will make +/-54Vdc.
This would fit perfectly what comes to the voltages +-78Vdc, +/-12Vdc, +15V driver (referenced to negative rail voltage)
Will get back to the other questions.
BR, Sami
Last edited:
Perhaps I am missing the point here but the amp modules run on unregulated +/- rails.
I should have said UcD700 in post #42. I note the UcD400 has a "typical" main power supply voltage of +/- 75V. For the UcD400 VDR is 13V rather than 15V.
UCD400oem typical voltage is +/-57V, not +/-75
Probly I will change the driver voltage regulator to LM317, and you will be able to change this voltage by just 1 or two resistors. Or I could even place a trimmer pot there to make it real simple, although you have to be careful and adjust the pot prior to powering the ucd...
Indeed. I hate that one can't correct posts for typos.
I wouldn't use a pot. The voltage divider is simple enough. I just built a regulator based on the LM317 but with an SPX431 anchoring Vref and a J310 anchoring Aref. A few more parts but a leap in quality.
I wouldn't use a pot. The voltage divider is simple enough. I just built a regulator based on the LM317 but with an SPX431 anchoring Vref and a J310 anchoring Aref. A few more parts but a leap in quality.
Resistors R5-14? Although I note there's no cap after the R which was going to be another one of my questions.... 🙂
I recommend reading the pdf in the first post of the thread I linked to above. It's the resistor which does the snubbing.
(I'm sure it is obvious to you that doing a lot of work on the board in Eagle is poor-return work until you have completely finalised the schematic - as I have learnt from very recent experience! 🙂 )
Doesn't matter which side of a source-drain short the detector is on; current through it is the same either way since the FET and detector are in series. An alternate fault path does exist through the gates but the photocouplers are rated 4.5kV versus 2.1kV for the ACS712. In the unlikely event the couplers fail but the detector survives "primary" side placement of the detector would catch this whereas "secondary" side wouldn't. But, since a gate fault doesn't go through the load, a fuse blow should be sufficient.
That was very interesting reading! I have learned a lot...
Here is the next version (png, eagle)
* I have made the fixes suggested by doctordata and others
* Added real snubbers (anyway, I'm probably going to place these directly on the secondaries, but will make the board with places for the snubber components)
* added capacitors C41/C42
* Replaced the rectifiers on main PSU board
* Removed overcurrent protection
* Changed the PICAXE from SMD to DIL
* 2 parallel FETs instead of 4
..Improvements to follow later, but comments are welcome...
I will finalize the board when we are close to the ready with the schematics 😉
Last edited:
I would be interested in people's thoughts around this. Does it make a difference? One could even deploy a Jan Didden super regulator in each of these.
I would also appreciate anyone's thoughts on interpreting the NC400 datasheet with respect to the above (it doesn't appear to have the same flexibility as the OEM model). On pg 4 it calls for Vsig between 16V and 25V. There's also provision for an "external Driver supply voltage". Is the 16V mentioned as "typical" a typo - should 16V not be the minimum? Also, note 1 says
while note 2 says:
Can someone explain the meaning of this and why?
EDIT Actually on pg 5 I think we have clarification for at least my first question:
Interesting that this should be unregulated...
Last edited:
Thanks.
Check the polarity of D27 and D29?
What is driving the thinking re sequence of caps, their values and resistors in the main supply stage? I contrast it with setup in the Universal PSU V3 That is, no large cap after the resistor bank. (BTW that PSU PCB and the soft start module provide some good ideas for layout and connector selection.)
I've never heard of double regulation being a bad idea. Can you elaborate?
(Related to this, the option on the NC1200 to not have on-board regulation relates, I presume, to signal voltage supply. The NC400 calls for 16V<Vsig<25V. I presume the corollary of this is that the signal supply voltage is also regulated on-board?)
(Related to this, the option on the NC1200 to not have on-board regulation relates, I presume, to signal voltage supply. The NC400 calls for 16V<Vsig<25V. I presume the corollary of this is that the signal supply voltage is also regulated on-board?)
- Status
- Not open for further replies.