Dear All,
This is my first posting here, my question is if a 4:1 input transformer would work on the Pascal S PRO2 module.
The gain of the S PRO2 is 26db and it has a 2k2 input impedance.
My tube preamp has 20db gain from parallel 12au7, anode output, output impedance 2k5.
I found the following transformers: Lundahl LL1545 and Jensen JT 10 KB-D.
Would this work or will I loose too much gain because of the 4:1 transformers?
A alternative could be a output transformer in my preamp.
Any thoughts or experience about/with this?
Kind regards!
This is my first posting here, my question is if a 4:1 input transformer would work on the Pascal S PRO2 module.
The gain of the S PRO2 is 26db and it has a 2k2 input impedance.
My tube preamp has 20db gain from parallel 12au7, anode output, output impedance 2k5.
I found the following transformers: Lundahl LL1545 and Jensen JT 10 KB-D.
Would this work or will I loose too much gain because of the 4:1 transformers?
A alternative could be a output transformer in my preamp.
Any thoughts or experience about/with this?
Kind regards!
The 4:1 transformer will cause a loss of -20*log10(1/4) = 12dB.
The effective input impedance will by scaled up by the square of the ratio: 2.2K * 4^2 = 35.2K
There will be a voltage divider action between the preamp output impedance of 2.5K and the 35.2K resulting in an additional loss of -20*log10(35.2/(35.2+2.5)) = 0.6dB
Giving a total insertion loss of 12.6dB
The effective input impedance will by scaled up by the square of the ratio: 2.2K * 4^2 = 35.2K
There will be a voltage divider action between the preamp output impedance of 2.5K and the 35.2K resulting in an additional loss of -20*log10(35.2/(35.2+2.5)) = 0.6dB
Giving a total insertion loss of 12.6dB
These modules are not high end, the channels idle battling, due to a 20khz to 30khz difference in switching frequency, which strongly depends on phase shift shown at 400khz by the MC33079 low cost quad op-amp used. That 20khz to 30khz beating component is present in idle noise spectrum, and its harmonics and subharmonics too, far more than shown in datasheet, there is a peak of about 20dB at 10khz in noise floor that moves with low frequencies at low listening volumes. Also, 1st op-amp stage has 2.2k input impedance, but 2nd stage has 1k input impedance and a 100uF 6.3V bipolar electrolytic in series, this also leaves a signature in bass at low listening levels, as the capacitor used that way has high dissipation factor (lossy below 100hz).
Fortunately the tubes probably have higher noise floor, if 20dB higher than the amplifier, the peak around 10khz in noise floow will be completely covered. Transformers are lossy at LF too, having a similar weakening effect on bass.
The power supply is dirty too, it is advertised as "PFC" but the module does not qualify to be subject to PFC regulations, 1/8th power at 2x 8r is below 200W. In fact the SMPS has *negative* input impedance putting a lot of distortion into the line at high power demand, although power factor at low power is OK. The power supply has no low-power burst mode, so it also introduces a n*50hz spray of harmonics in noise floor, not as bad as the HF 10khz peak, probably humm and buzz in the tube circuits will cover this too. The ultimate amplifier SMPS. What comes after LLC? Single stage PFC.
Fortunately the tubes probably have higher noise floor, if 20dB higher than the amplifier, the peak around 10khz in noise floow will be completely covered. Transformers are lossy at LF too, having a similar weakening effect on bass.
The power supply is dirty too, it is advertised as "PFC" but the module does not qualify to be subject to PFC regulations, 1/8th power at 2x 8r is below 200W. In fact the SMPS has *negative* input impedance putting a lot of distortion into the line at high power demand, although power factor at low power is OK. The power supply has no low-power burst mode, so it also introduces a n*50hz spray of harmonics in noise floor, not as bad as the HF 10khz peak, probably humm and buzz in the tube circuits will cover this too. The ultimate amplifier SMPS. What comes after LLC? Single stage PFC.
Last edited:
Thank you both for your valuable information!
JohnnyInBris, it seems I might have enough gain to play with and worth trying a input transformer.
Eva, these modules may not be high end in terms of measurements but they are well regarded in various well known amplifiers, Rowland for instance.
It seems to me that adding more opamps in the signal path would not be a sound idea
as the sound is already "compromised" by the use of the MC33079.
Anything in front with a higher slew rate than 7.0 V/µs does not make sense if I understand this correctly.
JohnnyInBris, it seems I might have enough gain to play with and worth trying a input transformer.
Eva, these modules may not be high end in terms of measurements but they are well regarded in various well known amplifiers, Rowland for instance.
It seems to me that adding more opamps in the signal path would not be a sound idea
as the sound is already "compromised" by the use of the MC33079.
Anything in front with a higher slew rate than 7.0 V/µs does not make sense if I understand this correctly.
The sound is compromised by the aggressive reduction of development time and part count. Apart from the dirty power supply, that kind of class D does not even use a comparator or a proper post-filter feedback path, it uses phase-shift of low cost op-amp for oscillation, it is mixed pre-filter and post-filter feedback, high damping is lost at high freq, and the logic input of IRS20957 is used as a comparator, switching frequency drifts substantially, after repairing one module damaged from input transient the offset between channels was no longer 30khz, but 10khz instead, so huge 30dB peak in noise floor at 10khz. Monacor (brand like Parts Express) ended up selling amplifiers with these modules too https://www.monacor.com/en-gb/monac...a-amplifiers/digital-pa-amplifiers/sta-1000d/ (and STA-2000D, and Powersoft Litemod is the same thing but with worse specs).
Damn, I no longer have a good amplifier. I believe what you say, but many listening test is in favour for the Pascal modules. Not all I know, but at least many. The modules are used in both very very expensive PA amplifier and high end hifi amplifiers as well as very cheap PA amplifiers. Monacor also uses Hypex modules, and they are even cheaper than the Pascal versions. I acutally thought that a mixed pre and postfilter as also seen in ICEpower was a good thing.
Pascal S-PRO2
Don't be to disheartened about this module, I am driving it now from a ALIENTEK D8 headphone output with good results so far.
I bypassed all the electronics of the STA 1000D input module, not difficult to do just remove 4 resistors and 2 coupling caps.
Careful though,by doing this no DC should be present from your source unless you leave the input board`s output coupling cap in place.
Be very patient, this module requires 500 to 1000 hours to burn in
Please read about it on this forum thread:
Class D = Trash? | Audiogon Discussion Forum
You will loose warranty by doing this unless it is done reversible
and invisible
when the need for repair arises...
I am not taking any responsibility for what you do to your amp!
Don't be to disheartened about this module, I am driving it now from a ALIENTEK D8 headphone output with good results so far.
I bypassed all the electronics of the STA 1000D input module, not difficult to do just remove 4 resistors and 2 coupling caps.
Careful though,by doing this no DC should be present from your source unless you leave the input board`s output coupling cap in place.
Be very patient, this module requires 500 to 1000 hours to burn in
Please read about it on this forum thread:
Class D = Trash? | Audiogon Discussion Forum
You will loose warranty by doing this unless it is done reversible
and invisiblewhen the need for repair arises...
I am not taking any responsibility for what you do to your amp!
The IRS20957 chips in these modules can become damaged due to transients by swapping input signal connectors (with no buffer/transformer) while the module is powered up, it happened to me. Design is not for direct signal connection to the module, but it does have a 100uF/1kohm HPF at input (2nd stage), so DC is filtered out.
Class D = Trash? Module manufacturers are happy to produce trash if it sells, and often overestimate the quality of whatever else is already on the market, fooled by the salesman of other manufacturers (the worst the Italian and the Chinese, these could sell trash with a 1um golden finish), I'm a class D designer and I've had arguments with manufacturers from time to time, me wanting to include more quality, manufacturers hesitant when facing the complexity of producing better circuits (few more parts, daughter PCBs, 1 or 2 more steps at factory to complete a batch of modules).
Class D = Trash? Module manufacturers are happy to produce trash if it sells, and often overestimate the quality of whatever else is already on the market, fooled by the salesman of other manufacturers (the worst the Italian and the Chinese, these could sell trash with a 1um golden finish), I'm a class D designer and I've had arguments with manufacturers from time to time, me wanting to include more quality, manufacturers hesitant when facing the complexity of producing better circuits (few more parts, daughter PCBs, 1 or 2 more steps at factory to complete a batch of modules).
Hi EVA,
Could you make your own project? - like Lars Clausen - I don't think he can make a living with New Class D modules, but I bet it's fun for him
There is a little group in Denmark, also trying to designing the ultimate class D right now. It's always fun follow such projects
Could you make your own project? - like Lars Clausen - I don't think he can make a living with New Class D modules
, but I bet it's fun for himThere is a little group in Denmark, also trying to designing the ultimate class D right now. It's always fun follow such projects
Last edited:
Hi EVA,
Could you make your own project? - like Lars Clausen - I don't think he can make a living with New Class D modules
There is a little group in Denmark, also trying to designing the ultimate class D right now. It's always fun follow such projects
I'm currently involved in a similar project with an US manufacturer. Clean and disciplined enough to be high end. Powerful and compact enough to be PA. Complete and function-rich to be stand-alone, but also supporting digital control interfacing. Simplified and functionally optimized enough o be affordable. Progress is slow, income is low, emotional feedback from customer is only satisfactory occasionally, considering that the stuff I'm doing is probably going to be up to date for 10~20 years.
The society in the country where I live is seldom supportive either. If I was reliably offered a bedroom, a lab room and a piece of garden in another country I would abandon the absurd touristic place where I live.
I'm not going to fail on the customer, but if they fail on me I will contact the little group you suggest. They probably have the ideas clear for class D section, but I have a nice evolution of the PSU used in Pascal S-PRO2 and Powersoft Litemod, featuring true PFC and shutdown-less hiccup-less amplifier overload, record full load efficiency thanks to IGBT, and record low load loss thanks to smart power management.
Last edited:
Calculating Output Zobel/ Input filter for power amp module
I would like to introduce a filter stage between the Jensen JT-10-KB-D and
the Pascal S-Pro 2 module to limit bandwidth reaching the S-Pro 2.
How do I calculate this, I have a input resistance of 2k2 on the module, a DCR from the transformer secondary of 91 Ohm in parallel with the 2k2.
Can I just put a cap parallel to this or would this create a problem because
of the inductance of the secondary of the transformer?
Would a Zobel be a better solution?
Someone from JRDG mentioned that to get the best out of a class D amplifier
it would be good to limit the incoming bandwidth to the module.
Please advice,
Thanks
I would like to introduce a filter stage between the Jensen JT-10-KB-D and
the Pascal S-Pro 2 module to limit bandwidth reaching the S-Pro 2.
How do I calculate this, I have a input resistance of 2k2 on the module, a DCR from the transformer secondary of 91 Ohm in parallel with the 2k2.
Can I just put a cap parallel to this or would this create a problem because
of the inductance of the secondary of the transformer?
Would a Zobel be a better solution?
Someone from JRDG mentioned that to get the best out of a class D amplifier
it would be good to limit the incoming bandwidth to the module.
Please advice,
Thanks
