Impedance Matching


Would you even consider running a tubed preamp with an output impedance of approx 2,500 ohms into a lower impedance solid state amp, say if the amp had a value of 15,000 ohms?
stickman451
I would certainly consider it- you won't damage anything, it just may not sound its best.

Could be you might get a loss of low frequencies with this combination.

But that is not guaranteed. The output impedance at 20Hz will tell you. If it is the same value then you have no worries. If it is a lot higher at 20Hz they you might have a problem.

In general, the output impedance should be 1/10th of that which it is driving.
Ralph, thanks for chiming in. I had brought this issue up in another of the OP's current threads. The preamp in question is a Hovland HP-100 with the moving coil option. Its output impedance, per JA's measurements in Stereophile, is 2.4K (presumably at 1 kHz), and 4.3K at 20 Hz.

The power amp he is considering is a Bryston 28BSST2, which is specified as having an input impedance of either 15K or 16.5K, depending on what document is looked at. He is using Maggie 20.7 speakers.

The opinion I expressed is that while that combo may work reasonably well with many recordings, especially those having limited deep bass content, I see no point in introducing what is at best a known marginality into the system, when other alternatives are available that would avoid the issue.
The output impedance at 20Hz will tell you. If it is the same value then you have no worries. If it is a lot higher at 20Hz they you might have a problem.
This is an excellent point that I think warrants emphasizing. What matters most is not the relation between amp input impedance and the nominal output impedance of the preamp, but the relation between amp input impedance and the range of VARIATION of preamp output impedance as a function of frequency. That is why, for example, some people report that resistance-based passive preamps, having VERY high output impedances, still work fairly well into amps having input impedances such that the 10x guideline does not come close to being met. The output impedance of that kind of preamp will have very little variation as a function of frequency.

Regards,
-- Al
If you can audition the amp with the preamp then sure. But other than that, I would not go that route, especially if there was a potential for losing money in a re-sale on the used market.
Unfortunately I cannot 'try before I buy' but the price is very, very, good and hard to pass up. Magnepan voices and runs their speakers with Bryston gear because (1) it sounds good to them, (2) the Brystons NEVER break driving the DIFFICULT Maggies.

At the bargain price I almost wonder if it would be worth picking up a good SS pre to drive them and just keeping the Hovland too!
At the bargain price I almost wonder if it would be worth picking up a good SS pre to drive them and just keeping the Hovland too!
Another possible approach, although it would result in having an additional active stage in the signal path, would be to buy the Bryston, and if the Bryston/Hovland combo proves to be unsatisfactory purchasing and installing a separate active buffer stage in between them.

I'm not sure if the Burson AB-160 buffer is still being made, but if you can find one new or used that would probably be a good choice. Also, A'gon member Bifwynne (Bruce) has had a custom buffer stage made for him at a modest price by Tom Tutay of Transition Audio Design, I believe in Florida. Tom is reported to be an outstanding person to work with, and Bruce reported excellent results with the buffer stage.

Either of those devices would provide a very high input impedance and a very low output impedance, just what you need.

Regards,
-- Al
Stickman ..., Al is correct. Tom Tutay built a custom impedance buffer for my subwoofer that sums the L and R channels and also ties into the balanced outputs of my linestage.

I have a caveat that is best addressed by Al and/or Ralph. My impedance buffer only ties in the subwoofer, which is self powered. There are no artifacts between my linestage and my main amplifier. I only use the sub to supplement the bottom octave or two (below 50 or 60 Hz) where my main speakers roll off.

So, perhaps Al and/or Ralph can speak to whether inserting an impedance buffer between the pre and amp could compromise the signal.

Bruce (bifwynne)
... perhaps Al and/or Ralph can speak to whether inserting an impedance buffer between the pre and amp could compromise the signal.
Certainly inserting an additional active stage into the signal path to correct an impedance incompatibility can invoke some tradeoffs, to a greater or lesser degree depending on the quality of the buffer.

But although nothing in the preamp itself would be changed, I would view doing this as in effect amounting to implementing a design modification to the preamp, with the modified "preamp" consisting of the combination of preamp + buffer stage, and a (short) pair of interconnecting cables between them. (It would most likely be best to make the cables on the input side of the buffer as short as possible, with those on its output side as long as necessary, so that the longer cables are driven by the device having low output impedance).

Viewed in that manner, it seems to me that even apart from resolution of the impedance incompatibility, the subjective effects of adding a buffer stage would stand a reasonable chance of being an improvement, rather a degradation, although either outcome is certainly possible.

In any event, the OP indicated in his other thread that the opportunity he had to purchase the Bryston amp went away, so the question may be moot.

Best regards,
-- Al
Post removed 
One of the big issues of impedance matching besides the frequency problems and other issues is the transfer of signal.

Using a very simplistic example.
Using only voltage you have a ratio of transfer. So going from 10k to 10k you would have 50% transfer. Going from 1k to 10k would give you a 1 to 10 transfer ratio and so on. Generally, so you don't have current loading issues and voltage transfer issues, it is best to have a low impedance source to a very high impedance input.

Anything else is tone control to an extent and that is why interconnects make a bigger difference on some systems then others.
Al, Got another technical question for you. How does an RCA to XLR adapter work? If you connect your high output impedace pre via single ended rca to a low impedance power amp and you use an xlr converter to go into the amp's xlr input, are you getting any benefit from interfacing with with the amps higher input impedance on its balanced in? (Often times it seems that big SS amps balanced-in has a higher specd input impedance)
Stickman, often (although not always) the number provided in specifications of balanced input impedances represents the sum of the input impedances of the two input signal "legs." An RCA-to-XLR adapter routes the signal on the center conductor of the RCA cable into one of those two legs (usually pin 2) while connecting the other one (usually pin 3) to ground (pin 1).

So in those cases the unbalanced output of the component providing the signal would see a load impedance that may be in the vicinity of half of the specified input impedance. Or, if not half, a load impedance that is substantially less than the specified balanced input impedance.

The Bryston amplifier you were considering appears to be an exception to that. While there are some inconsistencies between the input impedance values shown in the manual and the brochure, and the values that can be calculated from the schematic, it appears that they are defining balanced input impedance as being the input impedance of each leg. And that value is in the same general area, 15 to 20K, as the unbalanced input impedance. So the preamp would see a fairly similar load, and perhaps an identical load, with or without an adapter.

A separate question, apart from impedance considerations, would be if an amplifier having both balanced and unbalanced inputs would in itself sound better, the same, or worse using the balanced input with an adapter vs. the unbalanced input. Any of those three possibilities is conceivable, IMO, depending on the particular design.

Regards,
-- Al
Stickman451 hi, I have post on a similar thread some extensive tests that were done that should answer your question here;

http://forum.audiogon.com/cgi-bin/fr.pl?aamps&1387725945&read

Cheers George