Hiper
We have a PCB based on RDR-189. Eff 88%, Load regulation 24.18-24.15 no load 8A load. But ther is a 7KHZ oscillation on the resonating current measuring C ser current. The output is physically close to the input.
Comments
I can do that but the problem has been fixed by simply adjusting the output voltage from 24 V to 22V.
We are using a new transformer from Santronics that is like the RD-189 one but with no 12 V winding.
I have had a similar problem with another supply which is also similar but with a 60V output. It had the same problem which was fixed by lowering the output to 57 V.
Could the transformers simply need another turn on the output?
I sent this to Santronics on SNX-2458:
Also I measured the leakage inductance with the resonance method and I got 181 uH.
The spec says 110 uH.
I’m thinking that would be less coupling and lower output.
The transformer seems to work fine otherwise.
Is it as simple as adding turns to get a higher output?
As long as the leakage inductance is consistent that is OK.
Hi,
Leakage inductance in a transformer mainly depends on separation distance from primary to secondary. You can reduce the leakage inductance in an LLC based transformer by reducing the distance between primary and secondary winding. But at the same you have to look for the isolation requirements as well. Increasing the leakage inductance will increase the circulating currents and will reduce your efficiency.
The ratio of primary inductance to leakage inductance about 3 to 7 is an optimal choice in LLC based converters. With the original value of leakage inductance the ratio is 5 and now with the increased value of inductance the rato came down to 3. The ratio is with in the range but you will see some degradation in efficiency.
Output voltage of LLC converters is mainly depends on frequency of operation and expected frequency at nominal input voltage and full load is heavily influenced by turns ratio and and primary no. of turns. So adjusting the primary no. of turns will give the desired output voltage for a given frequency of operation.
I am ordering a transformer with 10% less primary windings. I have been doing some more testing on the current design. I have found the supply resonnat current is stable above resonance. But below resonance I see a discontinuous rise in the magnetising
I am ordering a transformer with 10% less primary turns. I have done some more testing on the existing design. Above resonace the circulating current is stable. Below resonance though there is a discontinuity in the magnetising current and the circulating current beconmes unstable with a several khz oscillation superimposed on it. Is that a symptom of a transformer that cannot support the output voltage? Normally the circuit should work above and below resonance, correct?
Here is a photograph of the scope screen of below resonance.
Hi,
The converter should work for above resonance and as well as for below resonance condition. I have got couple of waveforms from you. In one of your post the picture shows it is completely unstable for the below resonance condition and in the other post it shows a stable waveform for the below resonance condition.
Could you please specify under what conditions these two waveforms were taken?
Could you please capture the HB Voltage with respect to source pin?
In your previous post you said you were increasing a turn, but in your recent post I read your comment saying 10% lesser primary no. of turns.
Could you please share your PIxls file, so that I can review your design calculations?
Regards,
PI-NANO
In one of your post the picture shows it is completely unstable for the below resonance condition
The unstable picture is superimposed traces of below resonance
and in the other post it shows a stable waveform for the below resonance condition.
The stable waveform is one trace
Could you please specify under what conditions these two waveforms were taken?
There is an approximate 10 KHZ envelope oscillation of the circulating current. It varies with output voltage( e.g. frequency )The single trace blow resonance is just one trace. The envelope oscillation has numerous frequencies so the unstable waveform superimposes the different frequencies over each other.
Could you please capture the HB Voltage with respect to source pin?
Attached
In your previous post you said you were increasing a turn, but in your recent post I read your comment saying 10% lesser primary no. of turns.
Your response suggested removing primary turns is the better way to increase output voltage if I need to stay above resonance to eliminate the envelope oscillation.
Could you please share your PIxls file, so that I can review your design calculations?
Attached. I am not sure if the Santronics transformer is wound as suggested.I used their spec numbers for inductance even though I measured different.
Hi,
Please read through my previous post again. I did not suggested to reduce the turns. Reducing the turns will make the things even worse. I was going through your spread sheet, it says Lprimary=550uH and Lleakage=140uH. But I remember that your leakage was around 180uH. Please plug in the values whatever you have got from the transformer and if you get any error in the spread sheet, try increasing the no. of primary turns (The errors will go away) such that fnominal_predicted is slightly less than or equal to fser. I saw that you are having 15 secondary turns for each phase of main output and I see your primary no. of turns are 118. Please Increase the primary no. of turns to 140, so that fnominal_predicted is slightly less than or equal to fser. Increasing the no. of turns might increase your leakage inductance further. In that case increase your primary inductance (by reducing air gap) such that the ratio of primary inductance to leakage inductance will be greater than 3. With this your nominal frequency will be close to 80Khz. hopefully this should resolve your issue. If you want to have your nominal frequency close to 90Khz then try reducing the resonant capacitor value to 18nf from 22nf.
Regards,
PI-NANO
I have put my transformer in the RD-89 Eval board. It works fine. So thetransformer is not the problem.
Attached is a screen shot of HB vs circulating current on our board.
On RD-189 board with our transfomer the circulating current is perfectly symmetrical. Can you give me an idea of what the asymmetry in the circulating current is due to. I have adjusrted the voltage so the circulating current is just at resonance. This where the instabiility starts.
Proper Attachment
Proper Attachment
Hi,
Your schematic shows R290 as 27.4K, where as the spread sheet calculation is 19K. Would you mind trying it out.
Regards,
PI-NANO
That change was a suggestion from PI to allow hysteretic mode at no load.
I found the problem, it works good now. I had substituted an opto coupler with too low a bandwidth.
I am testing startup. There is a long delay before Vcomp comes up under load. . Under no load it comes up in a few milliseconds which is like the RDR-189 but loaded it takes several seconds. I have replaced all the components in that area and the PLC-810. Attached are two screen shots: at no load and 100W load. Shown is Vcomp ( 100 mv/div is 1 v/div)-yellow, input AC current -blue, V bulk - pink.
Noload and 100W labels are reversed on attachments.
The problem was ISP reference ground was not connected directly to the ISP sense resistor.
Hi,
Could you please send PCB layout and schematic files of your board.
Regards,
PI-NANO.