That has to do with the overall circuit design. If your phono allows for its power ground to be at mains earthed chassis potential without a problem the Ubib has no saying in it. Good thing is it has an integrated current limiter so in case of something is done wrong at the load's side it will prevent the fault current to reach full short value.
Thanks again Salas for your rapid response.
I guess only the manufacturer, Project, will be able to answer that question about 0V and earth.
Interestingly, the Project Phono preamp came with a SMP power brick with the earth pin of an UK mains plug made of plastic meaning there is no earth connection so the probable answer is 0V shouldn't be at earth potential.
I guess only the manufacturer, Project, will be able to answer that question about 0V and earth.
Interestingly, the Project Phono preamp came with a SMP power brick with the earth pin of an UK mains plug made of plastic meaning there is no earth connection so the probable answer is 0V shouldn't be at earth potential.
Phonos usually appreciate a metal housing at 0V power ground potential regarding hum shielding. That does not mean that they all like that same housing at mains protective earth also. If the PSU housing is a separate box then you can earth that and connect power ground to the main box only without wiring the two chassis together. If its one chassis, you can connect protective earth to it also, but safe lift the PCB's ground from it with a bridge of diodes. Maybe you have seen such safe lift schemes in power amps to prevent ground loops but to also conduct and provide a fault current path in case of a hard failure.
Yes, I do have the PSU in a separate case and have connected the case to earth for safety purposes. However the earth is not connected to the ultrabib.
When you say I can connect the power ground to the preamp without connecting the cases together, does 'power ground' mean 0 on the output terminal of the ultrabib?
My reasoning for connecting the preamp to earth is for shielding as putting my hand on the preamp reduces humming a bit.
When you say I can connect the power ground to the preamp without connecting the cases together, does 'power ground' mean 0 on the output terminal of the ultrabib?
My reasoning for connecting the preamp to earth is for shielding as putting my hand on the preamp reduces humming a bit.
It means 0 on the Ubib's output or some connected to it 0 point on the phono's PCB. Usually such a point at the input stage is better for wiring to chassis regarding hum results. Sometimes they use a 100 Ohm resistor paralleled with a 0.1uF film or ceramic capacitor instead of a straight wire. In both cases keeping it short to chassis helps dumping RFI to it. See what works better in your case i.e just a wire or R//C.
1.5A for an op-amp board (?) sounds much excessive. Even when discrete and when there is a controller. What is its real mA consumption for the +/-15V rails, can you measure? You wouldn't want to set the CC way high with R1 for no reason because there will be much heat and possible stability problems.
About parts there's nothing too special to say. Nichicon KZ for C2 and Nichicon ES for C3 work well. You may use 10000uF for C1 if the consumption is really beyond 1A and demands high CC setting. Some Nichicon or Panasonic. Use good brand 50 ppm resistors, fast & soft TO-220 bridge diodes like MSRF860G or VS-HFA04TB60PBF, and don't forget to put the negative section's Q1 in reverse orientation than what the print shows on the green test board. Also choose 10k VR1 for more turns in the voltage area of interest if you don't plan any higher than 25V rail application.
About parts there's nothing too special to say. Nichicon KZ for C2 and Nichicon ES for C3 work well. You may use 10000uF for C1 if the consumption is really beyond 1A and demands high CC setting. Some Nichicon or Panasonic. Use good brand 50 ppm resistors, fast & soft TO-220 bridge diodes like MSRF860G or VS-HFA04TB60PBF, and don't forget to put the negative section's Q1 in reverse orientation than what the print shows on the green test board. Also choose 10k VR1 for more turns in the voltage area of interest if you don't plan any higher than 25V rail application.
Finally finished second board and wiring so now I can listen to my new rig.
I modified a paradise board so inputs are now jfet and no need for ips current nulling.
jfet ips has less gain so I modded the main mirror so it now has gain.
Result is 63dB at 1khz
The ultrabib is super stable and noiseless and I have one pos and one neg to power both riaa boards.
The result is an exceptionally silent preamp (no noise no hum) with an energic sound with really outstanding bass but this time with quite clear and extended highs as well.
I modified a paradise board so inputs are now jfet and no need for ips current nulling.
jfet ips has less gain so I modded the main mirror so it now has gain.
Result is 63dB at 1khz
The ultrabib is super stable and noiseless and I have one pos and one neg to power both riaa boards.
The result is an exceptionally silent preamp (no noise no hum) with an energic sound with really outstanding bass but this time with quite clear and extended highs as well.
Attachments
Each phono board spends 70mA pos and neg so I need at least 140mA from each shunt.
Setting current output in the shunts for 180mA with 3.3ohm R1 does not work.... negative shunt does not light up.
Setting shunt current to 220mA (R1 = 2.7ohm) works perfectly so I have a 70mA margin on the voltage mosfet.
Setting current output in the shunts for 180mA with 3.3ohm R1 does not work.... negative shunt does not light up.
Setting shunt current to 220mA (R1 = 2.7ohm) works perfectly so I have a 70mA margin on the voltage mosfet.
Very nice results and build. Also an interesting Paradise phono mod. Congratulations.
There's possibly some inrush current flow for the Paradise's input stage large value decoupling electrolytic capacitors and/or to its mirrors low R high C rail filter cells so you needed more CC in the end than you initially thought of.
There's possibly some inrush current flow for the Paradise's input stage large value decoupling electrolytic capacitors and/or to its mirrors low R high C rail filter cells so you needed more CC in the end than you initially thought of.
Luckily the 2SK170s are GDS pin order and the BC337/327s are BCE so you replaced them right in the Paradise's PCB with no leg tricks
The Katana documentation ( Amazon Drive ) is not clear on the exact +/-15V power requirements, but there is one useful spec. The +/-15V is generated by an on-board switching DC-DC converter. That converter's power requirements are stated to be 5V @ 180mA. This hints to about 30mA of current draw per 15V rail.
But then to make things a bit confusing, there is also this spec: "5V-800mA load for OPAMP".
So it's best to ask Allo.
But in any case, the power consumption will be nowhere near 1.5A.
I would imagine if you will ever add another one full PSU, i.e. extending to dual mono Tx + dual mono UltraBiBs, to can possibly extract even more subjective performance out of that Paradise special mod version. Of course you had experienced the original circuit with its own dual mono on-board PSUs before and any deterioration due to single external shared PSU would have been alarmingly obvious.
So no noise and good sound on a 63dB gain phono shows that using a single shared UltraBiB is still adequate for very good cooperation with such sensitive high gain type of circuitry.
@ Dimdim
As I suspected, there must be some confusion going on with that Katana stack's third floor board consumption spec. Thanks for looking into it. Anhton82 must investigate his real mA needs for correct CC setting a bit further now.
Of course a power test with some dual bench supply that directly displays the consumption or with LM317-337/LM7815-7915 or other handy diy PSU plus a DMM in series current mode between it and each rail input of the loading circuit will answer the steady/peak mA question without a doubt.
As I suspected, there must be some confusion going on with that Katana stack's third floor board consumption spec. Thanks for looking into it. Anhton82 must investigate his real mA needs for correct CC setting a bit further now.
Of course a power test with some dual bench supply that directly displays the consumption or with LM317-337/LM7815-7915 or other handy diy PSU plus a DMM in series current mode between it and each rail input of the loading circuit will answer the steady/peak mA question without a doubt.
Attachments
I will build a second set of ultrabib (in it's green form) and add another TX in the near future.
Right now I will concentrate in listening to better evaluate it's strenghts and possible drawbacks.
I am really amazed with the bass performance.....
