Topping D10s + E1DA Cosmos ADC as a measuring system

[pma]

This combo make an excellent low cost measuring system, enabling to measure harmonic distortion components down to -130dBr, at about 300 EUR total price. I use the D10s with Olimex USB isolator to avoid signal ground loops.

Below please see the measurement of my discrete preamplifier Dispre 2 - JFET.

Dispre 2 - JFET english

The result is better than that of many older opamps, including OPA134 e.g.

To measure power amplifiers, I have built input dividers to protect E1DA ADC input, I do not want to use the tiny TRRS 2.5mm connector, also for the reason of possible short when moving the connector.

 

[mdsimon2]

Would you mind sharing the design of your input dividers? Just curious what you find to be the best for the relatively low input impedance of the Cosmos.

Thanks,
Michael

 

[pma]

Would you mind sharing the design of your input dividers? Just curious what you find to be the best for the relatively low input impedance of the Cosmos.

No problem, I use usual resistor dividers, both SE and balanced configuration, at low impedance (not to add resistor noise), like 1k / 100R and appropriate switch setting on E1DA.

The SE divider + 2.7V EDA setting adds 13x attenuation, so the max. input is 35.1V. Simple math, you can make it as you like.

SE divider looks like this

The Dispre 2 - JFET preamp was measured directly, without a divider, in a setup as below:

 

[mdsimon2]

No problem, I use usual resistor dividers, both SE and balanced configuration, at low impedance (not to add resistor noise), like 1k / 100R and appropriate switch setting on E1DA.

View attachment 211407

The SE divider + 2.7V EDA setting adds 13x attenuation, so the max. input is 35.1V. Simple math, you can make it as you like.

SE divider looks like this
View attachment 211408

The Dispre 2 - JFET preamp was measured directly, without a divider, in a setup as below:
View attachment 211409

Thanks for sharing, your resistor values are much lower than the what I have been using, I will give that a try.

I also really like the shielded box, mine have been loose and are sensitive to positioning / environment.

Michael

 

[pma]

Thanks for sharing, your resistor values are much lower than the what I have been using, I will give that a try.

Because the input impedances of the E1DA ADC are low, the resistors have to have low values as well. This is fine when measuring power amps or preamps with high output current like my D2-JFET, but it is not so good for many usual commercial preamps that do not have capability to drive 600 ohm load. Opamp buffer would be helpful here, but it must have:

- very low noise
- very low distortion
- ability to drive 500ohm - 1kohm load without increasing distortion

I tried my Audio Buffer (100k input impedance, 50ohm output impedance, 250mA output current) but it degrades the D10s/E1DA parameters, as you can see below. In fact my D2-JFET preamp works better, but has a bit higher noise. See the buffer 500ohm plot below. 45ohm plot is with additional 50ohm terminating resistor.

 

[pma]

I have bought a simple XLR splitter and configured E1DA Cosmos ADC into mono mode with 2 channels driven in parallel. This is the Cosmos configuration to get highest dynamic range. I have re-measured Topping D10s with this Cosmos configuration. Cosmos switches are set to 2.7V input voltage and with 2 channels in parallel this makes very low input impedance of 250 ohm. D10s, however, is able to drive it very well. As a reminder, D10s has output impedance of 20 ohm. So please take into account the voltage divider 250/(250+20) ohm which reduce 0dBFS output voltage of D10s to 1.76V.

Configuration of D10s + Cosmos loopback. Please note the USB isolator that is inserted in the USB link to D10s. The use is mandatory.

 

[pkane]

I have bought a simple XLR splitter and configured E1DA Cosmos ADC into mono mode with 2 channels driven in parallel. This is the Cosmos configuration to get highest dynamic range. I have re-measured Topping D10s with this Cosmos configuration. Cosmos switches are set to 2.7V input voltage and with 2 channels in parallel this makes very low input impedance of 250 ohm. D10s, however, is able to drive it very well. As a reminder, D10s has output impedance of 20 ohm. So please take into account the voltage divider 250/(250+20) ohm which reduce 0dBFS output voltage of D10s to 1.76V.

View attachment 211770


Configuration of D10s + Cosmos loopback. Please note the USB isolator that is inserted in the USB link to D10s. The use is mandatory.
View attachment 211771

Hi Pavel, that's a really excellent performance for such an inexpensive measurement rig!

What USB isolator are you using? I've been using Intona for my loopback measurements, but I see some much less expensive iso's becoming available with support for high speed USB.

 

[pma]

Hi Pavel, that's a really excellent performance for such an inexpensive measurement rig!

What USB isolator are you using? I've been using Intona for my loopback measurements, but I see some much less expensive iso's becoming available with support for high speed USB.

Hi Paul, this one

USB-ISO

Full speed USB isolator 1000VDC for protecting your PC or laptop while connecting USB devices wired to high voltages CE certified

www.olimex.com

and it is a real bargain!

 

[Veri]

Hi Paul, this one

USB-ISO

Full speed USB isolator 1000VDC for protecting your PC or laptop while connecting USB devices wired to high voltages CE certified

www.olimex.com

and it is a real bargain!

Never heard of this little thing Cheers.
Also very intriguing topic, impeccable performance for what it all costs!!

 

[pkane]

Hi Paul, this one

USB-ISO

Full speed USB isolator 1000VDC for protecting your PC or laptop while connecting USB devices wired to high voltages CE certified

www.olimex.com

and it is a real bargain!

Wow! That really is a bargain! But it appears to be only full-speed, so may not work with some DACs and at higher sampling rates.

 

[pma]

Never heard of this little thing Cheers.
Also very intriguing topic, impeccable performance for what it all costs!!

Yes, it just needs a bit qualified operator as you have to think about dividers, ranges, noises etc., but for an enthusiast DIY hobbyist with some el. education or qualification it should not be a problem. I spent 127 EUR + 233 EUR = 360 EUR total including VAT and shipping costs, for D10s + Cosmos, from Audiophonics.

 

[pma]

THD and THD+N (of this combo) vs. level, related to DAC dBFS output, at 1kHz:

 

[linear diver]

I am such a person! I just got my Cosmos and D10s. I hooked them up first thing this morning and here are my 'hello world' first measurements. Unfortunately mine do not look as good as pma's

My physical set-up is similar, BUT I don't have a USB isolator. Are the big spikes below 1k due to USB issues?

 

[pma]

My physical set-up is similar, BUT I don't have a USB isolator. Are the big spikes below 1k due to USB issues?

Welcome to the club and please believe me, I am quite experienced in finding and fixing EMI issues and gnd loop issues . Yes, your spikes and especially noise floor modulation and rise are a result of USB ground loop and a device with SE output.
Please also check this image

https://www.audiosciencereview.com/forum/index.php?attachments/e1da_dividers-png.211407/

carefully, the RCA - XLR cable between D10s and Cosmos must be shielded twist pair where the -sig wire from RCA body to XLR pin 3 serves to subtract interference voltage across the cable shield due to ground currents.

For any serious work, you will need the USB isolator. I gave a link and that unit works with Topping D10s.
You could also see the USB ground issues even in the @amirm reviews like explained here

Cambridge Audio CXA81 Review (Sample 2)

This is a follow up to my first test of the Cambridge Audio CXA81. Results there showed fair bit of noise which was uncharacteristic of other products I have tested from the company. There were a lot of speculation as to the source of the problem but no data to back any. Fortunately company...

www.audiosciencereview.com

so please take it serious. It is a learning process that is never ending.

 

[pkane]

I am such a person! I just got my Cosmos and D10s. I hooked them up first thing this morning and here are my 'hello world' first measurements. Unfortunately mine do not look as good as pma's

My physical set-up is similar, BUT I don't have a USB isolator. Are the big spikes below 1k due to USB issues?

This does look more like a ground loop/power supply noise leaking into it. Try using a laptop running on batteries to drive the ADC if you can. Or, put in a USB isolator.

 

[pma]

This does look more like a ground loop/power supply noise leaking into it. Try using a laptop running on batteries to drive the ADC if you can. Or, put in a USB isolator.

I really do not think so. If he runs both devices from the same PC, then there is no other loop than via USB cables (and then analog cable of course). Both devices Topping D10s and E1DA Cosmos are powered from USB ports, no other tracks to ground. The PC may be even grounded or not, there is no difference. The loop goes this way: USB1 > Topping D10s > analog link cable > E1DA ADC > USB2. USB1 has same ground as USB2, so here is the loop and here the loop has to be interrupted by the USB isolator. Nowhere else. I really do know what I speak about. If he uses 2 PCs rather than one, then the situation is more complicated and has more solutions.

 

[pkane]

I really do not think so. If he runs both devices from the same PC, there is no other loop than via USB cables (and then analog cable of course). Both devices Topping D10s and E1DA Cosmos are powered from USB ports, no other tracks to ground. The PC may be even grounded or not, there is no difference. The loop goes this way: USB1 > Topping D10s > analog link cable > E1DA ADC > USB2. USB1 has same ground as USB2, so here is the loop and here the loop has to be interrupted by the USB isolator. Nowhere else.

This is why I use two PCs: one driving the DAC the other (battery-powered laptop) -- the ADC.

 

[pma]

My physical set-up is similar, BUT I don't have a USB isolator. Are the big spikes below 1k due to USB issues?

Welcome to the club and please believe me, I am quite experienced in finding and fixing EMI issues and gnd loop issues . Yes, your spikes and especially noise floor modulation and rise are a result of USB ground loop and a device with SE output.

This does look more like a ground loop/power supply noise leaking into it.

The loop goes this way: USB1 > Topping D10s > analog link cable > E1DA ADC > USB2. USB1 has same ground as USB2, so here is the loop and here the loop has to be interrupted by the USB isolator.

This is why I use two PCs: one driving the DAC the other (battery-powered laptop) -- the ADC.

Guys, please let me refute any doubts. Below is a measurement of a preamplifier module in the D10s/E1DA loop. The preamp is powered from a small 2x15V class II power supply which can be considered isolated (100pF primary/secondary). D10s and E1DA from the same PC. The only loop is through USB and this is the nastiest possible loop. D10s to preamp is SE cable, preamp to E1DA pseudo-balanced by RCA - XLR 3 wire cable. Below are the results with and without the Olimex USB isolator in the D10s USB link.

The USB loop is the common mistake made even by quite experienced users. It is a silly issue which must be avoided by using the USB isolator, in case of using such combo like D10s+E1DA for measurements.

@pkane : Paul, how would I synchronize REW from 2 PCs? I need REW to make THD and THD+N vs. level and frequency measurements.

 

[pma]

Topping D10s (with USB-ISO) to E1DA Cosmos ADC (2.7V range) residual noise. Mono mode, 2channel in parallel, connected as shown in

https://www.audiosciencereview.com/forum/index.php?attachments/d10s-e1dacosmosadcmono-jpg.211771/

 

[mdsimon2]

Thank you for sharing the noise measurement.

Am I correct in saying that for a noise measurement like this you want to calibrate REW to consider attenuation from the voltage divider created by the DAC output impedance / ADC input impedance? For example play a 0 dBFS tone from the D10S, measure output voltage with a DMM, then calibrate REW such that it also reads what the DMM does (even though we know from your previous measurements the voltage divider is actually reducing output voltage to 1.76 V).

It seems like if you do not do this the voltage divider will show a reduction in noise that will not be present when using traditional high impedance inputs.

Michael