Preamp ramblings: Active, Passive, TVCs, Birds, And Bees

[bazelio]

Well we don't have a pot in a TVC or AVC passive. So maybe your concern is moot. Though I'm not entirely sure I understand your concern. In a TVC or AVC, it's just directly wired to the input and the output through a switch on the input side for source selection.

As far as user experience in actives vs passives, I suspect most users simply prefer the coloration of high end tube preamps and even some solid state to the transparent nature of passive inductive volume controls. I suspect it has far less to do with TVCs/AVCs not being a good fit objectively.

 

[Tortuga Audio]

As TomC's measurements reveal, the LDR's input cap varies non-linearly and creates THD within the audio frequency range that approaches 10% (!!) as you adjust the volume. You worry about different sound at different volume levels?!? Nothing I've ever seen can touch the LDR's ability to achieve this.

Have you heard of "The Truth" preamp? I've been playing music through and measuring one of these recently. It's using a photo cell based attenuator also. Arthur Salvatore waxes poetic about it here.

Those measurements are a testament to how irrelevant measurements can be (beyond a certain point) when compared to the subjective experience of actually listening and experiencing the sound with your own ears. Meanwhile we are 2 generations down the road since those measurements were made....by a competitor no less.

I've several customers who previously bought/owned/listened to "The Truth". In once instance a customer bought one of our LDR passive preamp, sold it and bought The Truth, listened to it for a while, then sold The Truth and bought our passive preamp again.

Anyway, my 2 cents.

Best,
Morten

 

[Empirical Audio]

The devil is in the details when it comes to passive attenuators.

The passive resistive attenuator can be excellent, however in order to be excellent it usually must be located at the destination amp, not at the source, and usually requires very low capacitance cables to avoid HF roll-off.

The passive transformer attenuator is less sensitive to these effects, but also changes the impedance that the source sees. The main problem with transformer attenuators is the quality of the transformer and whether it has saturation effects. Good multi-tap transformers can be had, but they are expensive, usually $1K each. I'll pick a good transformer every time over a resistive attenuator.

Steve N.

 

[Tortuga Audio]

The passive resistive attenuator can be excellent, however in order to be excellent it usually must be located at the destination amp, not at the source, and usually requires very low capacitance cables to avoid HF roll-off.

This is often raised as a potential issue with resistive passive attenuation but as a practical matter we rarely see this as a primary constraint since as long as the IC run between the preamp and amp is kept to say 6-8 feet or less the filtering knee is kept well above 20 kHz. For most home audio scenarios this is quite manageable although there will always be exceptions. We once had a reviewer run our passive preamp in his system that had 25 foot long IC's to his amps. Told him this would probably be a issue. Guy did the review anyway and had no complaints. I'd argue he had hearing issues.

 

[analogsa]

The passive resistive attenuator can be excellent, however in order to be excellent it usually must be located at the destination amp, not at the source, and usually requires very low capacitance cables to avoid HF

If it is located at the destination, then the cable capacitance becomes irrelevant, doesn't it?

And in any case, the standard engineering objection to passives does not address their audible issues which seldom if ever concern the high frequencies.

 

[Empirical Audio]

If it is located at the destination, then the cable capacitance becomes irrelevant, doesn't it?

Correct.

And in any case, the standard engineering objection to passives does not address their audible issues which seldom if ever concern the high frequencies.

My engineering design is always a combination of measurement, analysis and listening tests. Classical measurements have limitations. Ear/brain evaluations by the right person do not.

Steve N.

 

[bazelio]

The devil is in the details when it comes to passive attenuators.
...
The passive transformer attenuator is less sensitive to these effects, but also changes the impedance that the source sees.

Small changes in DCR as you change taps and then the load seen by the source will get lighter as you increase in frequency from 20Hz on up. So as long as the transformer (or in my case autoformer) is gapped for a reasonable inductive reactance at 20 Hz, I'd expect no problems (e.g. low frequency response will be flat) with a reasonable soure. Mine is around 15kOhm at 20 Hz. I've used it successfully with a variety of phono stages save for a few poorly designed (IMHO) OTL units with extremely high output impedance where bass rolloff was noticeable.

The main problem with transformer attenuators is the quality of the transformer and whether it has saturation effects. Good multi-tap transformers can be had, but they are expensive, usually $1K each. I'll pick a good transformer every time over a resistive attenuator.

In a purely passive line stage, transformer/autoformer is the way to go. Though, I see no advantage to using a transformer/autoformer in an active line stage where the output is buffered. And I also see no reason to use an LDR in pretty much any situation... but that's just me.

 

[bazelio]

I've several customers who previously bought/owned/listened to "The Truth". In once instance a customer bought one of our LDR passive preamp, sold it and bought The Truth, listened to it for a while, then sold The Truth and bought our passive preamp again.

Curiosity got the best of me a while back and I bought a used "The Truth". I now refer to it as "The Lie" after reselling it. Its devices drifted something fierce and I found it nearly unusable from the standpoint of needing to tweak channel balance on a day to day basis, which I have no desire to do. Its sound quality, though, I must admit was really not bad. Better than I expected for sure. I'd flat out say it sounded good.

 

[Tortuga Audio]

Curiosity got the best of me a while back and I bought a used "The Truth". I now refer to it as "The Lie" after reselling it. Its devices drifted something fierce and I found it nearly unusable from the standpoint of needing to tweak channel balance on a day to day basis, which I have no desire to do. Its sound quality, though, I must admit was really not bad. Better than I expected for sure. I'd flat out say it sounded good.

My understanding (and it's just only that, since I've not taken one apart and verified first hand) is that the truth does use an LED plus photoresistor (an LDR by any other name) for volume control. LDRs are notoriously challenging to maintain channel balance over time and much effort usually goes into trying to find matched pairs (and matched pairs of pairs - 4 total) for stereo attenuation and 8 LDRs for balanced attenuation. When we started we pursued matched LDRs like everyone else. We eventually embedded a measurement circuit and algorithm into our LDR preamps so all of that matching happens under the hood. Just plug in any set of LDRs, start an automatic calibration cycle (takes ~10 minutes) to measure their performance, and then use that empirical derived and data to force the LDRs to behave identically. Arguably a hard way around to accomplish volume control but it's all about the sound.

 

[bazelio]

I opened the one I had. It's two metal can opamps, one on either side of of the LDR per channel. Unity gain with some loss. Off the shelf switching supply. It looked like a garage science project. The ooamps were unidentifiable as he had epoxied the cans.

EDIT: At least, they looked like opamps.

 

[Tortuga Audio]

I opened the one I had. It's two metal can opamps, one on either side of of the LDR per channel. Unity gain with some loss. Off the shelf switching supply. It looked like a garage science project. The ooamps were unidentifiable as he had epoxied the cans.

EDIT: At least, they looked like opamps.

If only because we've done quite a bit of work with LDRs and buffer stages, my guess is they're probably discrete mosfet or jfet transistors configured as input and output buffers. We use an LSK170 JFET along with a constant current source as the core of our output buffer stage.

 

[bazelio]

I'd traced it with a DVM, and the pin connections to the TO-99 was a typical opamp pinout. Also the switcher was delivering +/- 12V or 18V as I recall. So it seemed like they were unity gain opamps. Also the quoted input impedance was too high for JFETs. Maybe MOSFETs, but I didn't seem like it. I also have pictures somewhere on my phone...

 

[Tortuga Audio]

I'd traced it with a DVM, and the pin connections to the TO-99 was a typical opamp pinout. Also the switcher was delivering +/- 12V or 18V as I recall. So it seemed like they were unity gain opamps. Also the quoted input impedance was too high for JFETs. Maybe MOSFETs, but I didn't seem like it. I also have pictures somewhere on my phone...

Based on that observation my guess is: unity gain opamp input buffer -> LDR attenuator (probably just a shunt if it's a single LDR per channel) -> unity gain opamp output buffer. I would describe that as a solid state unity gain active preamp. Odd that it comes up in the context of a passive preamp.

 

[bazelio]

It's not passive, and there was loss as it didn't quite make unity gain per my measurements. I also wasn't 100% sure it was an opamp - it might have been some other type of high speed buffer, as the inverting input pin (if it was an opamp) was floating. I've never breadboarded that, and I'm not sure what it does. I think the only reason it comes up in the context of passive preamp discussions (such as AVC/TVC) is that it has been claimed to offer even better transparency. (Not my finding)