Note that to measure the THD, a precision oscillator (Victor's 1 kHz) was used. This oscillator does add some spurs at multiples of 30 Hz, hence, they show in the THAT Driver output in this measurement as well.
Tom
IIRC, Victor's oscillators use LME49720 op amps which are sensitive to emissions from DECT phone base stations unless well shielded. Spurs at 10Hz intervals or something like that may be a symptom, again IIRC.
You recall wrong. The DECT frame rate is 100Hz and this is what is picked up.
Thank you. Could be something else of a similar nature though. Otherwise, what would be the physical source of such spurs in a linear oscillator?
The 30 Hz spurs were probably from my WiFi router. I'm not sure why it would emit 30 Hz spurs even with the WiFi is turned off. Unplugging it removes the spurs.
Tom
Tom
The 30 Hz spurs were probably from my WiFi router. I'm not sure why it would emit 30 Hz spurs even with the WiFi is turned off. Unplugging it removes the spurs.
Tom
Noise injected into the mains from the router?
I highly doubt that. More likely EMI radiated by the router getting into my test setup (which tends to be unshielded on my lab bench). The EMI could also be carried by the ethernet cables, which will happily act as antennae.
Tom
Tom
How would the SMPS-86 fare for something like 3xTHAT Receiver -> Input selector -> Volume control -> THAT Driver?
It would fare very well. Some may rightfully argue that it'd be a bit overkill, but it would certainly work well.
Tom
Tom
Hi Tom,
I'd like to use the THAT Driver for two reasons: as a preamp and also to get a differential output (my DAC has SE only) for the 2xMod86 that still sit in boards inside my closet
Few questions:
Thank you so much
I'd like to use the THAT Driver for two reasons: as a preamp and also to get a differential output (my DAC has SE only) for the 2xMod86 that still sit in boards inside my closet

Few questions:
- Could LME49710NA eventually be replaced w/ another opamp? Are you maybe working on a v2?
- Seen in posts #2 & #97 that volume pot should be placed between RCA and the Driver. Can I use a 100KBTx2 pot? I have an old broken Rotel RA1000 amp and would like to reuse the pot, for sentimental reasons.
- Would it be possible to tap a headphones out? How/Where?
Thank you so much
Could LME49710NA eventually be replaced w/ another opamp? Are you maybe working on a v2?
I'm currently at Rev. 2.3. 😉
I do have a couple of tubes of LME49710NA that I'll be happy to sell you a few of. They came directly from Mouser and are still in the ESD bag. If interested, drop me a line through email.
The MOD86 does not lend itself well to random component substitutions, so I suggest sticking with the LME49710NA.
Seen in posts #2 & #97 that volume pot should be placed between RCA and the Driver. Can I use a 100KBTx2 pot?
I'd recommend less than 100 kΩ for lower noise, but 100 kΩ will work. Just make sure it's an audio taper pot (i.e. intended for use as volume control). 10 kΩ if your source can drive it would be better for noise. That said, 47-100 kΩ are common values for volume pots.
Would it be possible to tap a headphones out? How/Where?
I'd connect the headphone amp at the output of the volume control. The THAT Driver won't drive a headphone directly. You'll need to add a headphone amp.
Tom
That Driver is already at 2.3? 😱 I thought it was 1.xI'm currently at Rev. 2.3. 😉
I will since I don't want to burn the Mod86's somehow and get a brick 😀I do have a couple of tubes of LME49710NA that I'll be happy to sell you a few of. They came directly from Mouser and are still in the ESD bag. If interested, drop me a line through email.
The MOD86 does not lend itself well to random component substitutions, so I suggest sticking with the LME49710NA.
And if I get the That Driver, and there's no foreseable substitute for them, I will get an extra pair as well.
It's from an old Rotel RA1000 so yes, should be intended for that.I'd recommend less than 100 kΩ for lower noise, but 100 kΩ will work. Just make sure it's an audio taper pot (i.e. intended for use as volume control).
OK, I guess I'd need a switch at the output of the volume control to choose where the audio goes. Anything I need to know in that respect?I'd connect the headphone amp at the output of the volume control. The THAT Driver won't drive a headphone directly. You'll need to add a headphone amp.
Thank you again, as usual
That Driver is already at 2.3? 😱 I thought it was 1.x
The most current THAT Driver is Rev. 1.02.
I took this question,
I'd like to use the THAT Driver for two reasons: as a preamp and also to get a differential output (my DAC has SE only) for the 2xMod86 that still sit in boards inside my closet
[...]
Could LME49710NA eventually be replaced w/ another opamp? Are you maybe working on a v2?
to mean that you were asking about the Modulus-86 revision. Guess I misunderstood. My bad. The Modulus-86 went through an opamp change from LME49710NA (up to Rev. 2.0) -> LME49710HA (Rev. 2.1) -> LME49710NA (Rev. 2.2) -> LME49720NA (Rev. 2.3, which is the current rev.) all along with TI's waffling on whether to discontinue the LME49710.
The THAT Driver uses the LME49710NA and for those who'd like, I have a tube of them that I'm selling with the boards. I'll keep doing that as long as it's feasible.
Swapping out the opamp in the THAT Driver is no big deal. The trick is to find a good audio opamp, single, in DIP-8. I could redesign the board for the LME49720NA. That would be pretty easy to do.
That said, my inventory is getting pretty cluttered and managing it all is getting to be more of a distraction than it's worth in some cases. So at some point I need to sit down and look at the sales data and perhaps do some pruning of the slower sellers. I'll probably delay that until I run out of LME49710s, though - and that'll take a while.
I will since I don't want to burn the Mod86's somehow and get a brick 😀
You could always make an adapter that used one channel of the LME49720 in the socket for the LME49710. If neither is available, the OPA1688 may be adaptable for the MOD86. It'll require some compensation changes, but it's possible. The OPA1688 is a dual opamp ... in SMD.
And if I get the That Driver, and there's no foreseable substitute for them, I will get an extra pair as well.
The opamp in the THAT Driver is easier to exchange. It's just that the LME49710NA is the best (lowest THD) single opamp in an 8-pin DIP.
OK, I guess I'd need a switch at the output of the volume control to choose where the audio goes. Anything I need to know in that respect?
Use shielded wire. Use a switch (or relay) with gold points intended for small-signal switching. The switches by C&K tend to be pretty good, but there are many others. Keep the openings in the cable shield to a minimum.
Thank you again, as usual
You're welcome.
Tom
Last edited:
Hi again Tom,
I’m thinking about adding THAT driver as a “preamp” to generate the balanced output from the DAC that would go to a pair of Modulus86 v2.2. I think that gain would be too much for my small room though. Is there a way I can reduce the 20dB on the amp? I think noise wise it’d be better to attenuate between the THAT and the Modulus?
Thank you
I’m thinking about adding THAT driver as a “preamp” to generate the balanced output from the DAC that would go to a pair of Modulus86 v2.2. I think that gain would be too much for my small room though. Is there a way I can reduce the 20dB on the amp? I think noise wise it’d be better to attenuate between the THAT and the Modulus?
Thank you
The Modulus-86 has a gain of 20 dB already, so there's no action required.
The THAT Driver has a gain of 6 dB. There's no way to reduce that without adding an attenuator somewhere.
Tom
The THAT Driver has a gain of 6 dB. There's no way to reduce that without adding an attenuator somewhere.
Tom
The Modulus-86 has a gain of 20 dB already, so there's no action required.
The THAT Driver has a gain of 6 dB. There's no way to reduce that without adding an attenuator somewhere.
Tom
Thank you. And sorry I was not specific enough.
Where exactly is the SE to diff taking place? i am assuming after the LME49710 and before (or inside) the THAT1646. If that was he case, wouldn’t it be better to tap the potentiometer in between? According to the gain structure best practises (if I understood correctly) we should attenuate after some gain, in order to give more distance between signal and noise, resulting in less output noise.
Thank you again
Where exactly is the SE to diff taking place?
In the THAT1646.
If that was he case, wouldn’t it be better to tap the potentiometer in between? According to the gain structure best practises (if I understood correctly) we should attenuate after some gain, in order to give more distance between signal and noise, resulting in less output noise.
If that's what the article says, then the article is wrong. In the concept of gain structure, the low-noise techniques from the RF world are applied to audio. This means:
- Put the highest gain first.
- Avoid attenuators.
But that's me. The rest of the world has settled on 26+ dB gain in power amps and some perceive an amp where you need to turn the volume control past 10 o'clock to achieve ear-piercing SPLs as "weak", so there's that consideration as well.
Tom
Sorry I didn’t express correctly, and by no means did quote that article correctly. Instead of “we should attenuate after some gain” I meant “we should move some gain before the volume pot”. This matches what you describe as your preference as well (you’d want half the gain in the power amp but moving it to before the volume so you can have 15V there).In the THAT1646.
If that's what the article says, then the article is wrong. In the concept of gain structure, the low-noise techniques from the RF world are applied to audio. This means:
Attenuators add noise, hence, degrade SNR. In an audio reproduction system, you'll always need one attenuator: the volume control. I'd put as much gain as needed to drive the system to clipping (or slightly (6-10 dB) beyond) before the volume control and as little as necessary after the volume control. In the ideal world, I'd have 6-10 dB gain in the power amp and run 15 Vpeak swing from the volume control to the power amp at the highest volume setting.
- Put the highest gain first.
- Avoid attenuators.
But that's me. The rest of the world has settled on 26+ dB gain in power amps and some perceive an amp where you need to turn the volume control past 10 o'clock to achieve ear-piercing SPLs as "weak", so there's that consideration as well.
Tom
So, if we’re on the same page, now to my real scenario. THAT driver has 6dB gain. Modulus86 has 20dB gain. Volume pot is supposed to go at the RCA input in the THAT board, so... before ALL the gain. I will assume as well that all gain in the board is provided by the THAT1646, no gain from the opamps, right? If there is no place between the gain and the output to tap a vol pot (in theory this would be my desired option, if it didnt mess up with the impedances), I’d probably want another gain generator before the THAT board. My DAC will provide 1.2V so an extra gain is not needed but wouldn’t hurt either for such special songs/albums (you always mention Brothers in arms).
Thank you again for your patience. I’m learning baby steps 🙂
In the ideal world, I'd have 6-10 dB gain in the power amp and run 15 Vpeak swing from the volume control to the power amp at the highest volume setting.
But that's me.
I think there is a good argument to go even further and run the power amp at unity gain. But that is outside of this topic and needs beer!
I think there is a good argument to go even further and run the power amp at unity gain. But that is outside of this topic and needs beer!
I can open a bottle but the cool thing would be to share the conversation and beers at the same physical place, right? 😎
- Home
- Vendor's Bazaar
- THAT Driver :: A minimalistic differential driver / preamp with 0.000021 % THD