Booze Hounds Labs Phono any good?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
W.Jung and JL.Hood and B.Duncan are some names to google if you have a serious interest in HiFi electronics...

As it goes, I built a Ben Duncan-designed, passive EQ, op-amp-based, RIAA pre-amp (ADEQ), circuit published in Hi-Fi News, late 1980s. It certainly wasn't cheap though it was good; I used it for years.

After something went pfooof and because all the devices used are now obsolete, I replaced it a year or so ago with the BHL design. I used stock parts and a simple 24v PSU of my own plagiarising c/w remote transformer, C-R-C rectification and "tracking pre-regulation".

The new circuit sounds much better than the ADEQ though, as noted, a little more gain would be useful when driving some ADCs. For normal replay even using a passive (LDR . . .) VC and elderly amps with lowish input Z, it's fine.

Notice all the guys recommending this preamp to you are using subjective terms such as 'imaging' and improper/irrelevant use of technical terms. This should raise some red flags that you are being recommended some snake oil.

Every DIY audio project seems to have a Job who tells you it'll never work. This one does. I'm sure you can do better at higher cost but it's well worth the asking price. Incidentally, the LPs I've digitised using it typically sound much better than the CD reissue thereof.

While I'm about it, "red flag" and "snake oil" are, as technical terms go, as irrelevant as any and a bit childish with it.

HTH

Dave
 
Last edited:
If you start talking distortion, the location and harmonic will be more important than the level of it so often that it's not always an indicator to care about. Headroom is great, and matters a lot if said device has unpleasant clipping. But I can say despite the moderate headroom of the BHL, it sounds very pleasing to the ears with LDR's. It's impressive on how much rich tone you can get.
 
I know I might be causing trouble, but I can't resist.

I found out the hard way that headroom is really, really important in a phono stage. The music may sound clean until a pop or click comes along and hard-clips the second stage. If there's an RC network in between, blocking distortion can occur, further distorting the signal. The result is 'harshness' caused by the preamp constantly overloading on every click and pop.

So, headroom. Let's see if I understand anything about it...

Let's say your cartridge has the typical reference level of 5mV @ 5cm/S. Let's say the loudest passage (highest modulation in a groove) is +20dB higher than that nominal reference level. That's 10X the voltage out from the cartridge, so let's say that the preamp in question will have to handle 50mV peak inputs. That should be no problem.

Mr. Boozhound says that each stage of the preamp has 30X gain, which means we're talking 1.5V (!!) peak output from the first stage. An all-in-one-go passive EQ network has insertion loss of about -20dB, so here we have a 10X reduction in gain. So now we're at 0.15V out from the 1st stage, after the RIAA EQ, on a signal peak that is +20dB from nominal reference level.

*** If the 2nd stage can cleanly amplify more than 0.15V (150mV) peak at its input, then it should be OK as far as voltage headroom is concerned. ***

That begs the question, can that second 2SK170 in the Boozhound preamp accept a big blast of 500mV signal at its gate?

Yes or No.

PS - As a quick check:
30dB gain 1st stage + -20dB loss in EQ + 30dB gain in final stage = 40dB gain total.

Checks out.
--
 
This guy knows what he's on about ^ . The second 2SK170 can't take that 500mV without distortion as it can only swing about 11 volts or so into a high inpedance load if the right FET is chosen (which it most likely won't be) so 7 or 8V is more likely if you are driving a 22K load.

500mV x 30dB (second stage) gives you about 15V. Now some of you may not know this, but 15V is more than 11V (which is this preamp on a really good day if you've been given JFETs with optimal IDSS) so it'll overload and distort.
 
I just ran this circuit through LTSpice. I would post its schematic with annotations, except that it's a commercial product and I'd probably get in trouble.

What I get is that the gate to source voltage (analogous to 'grid bias' for a tube) is 130mV. In a triode, the grid bias defines the highest voltage that that stage can accept before grid current and gross distortion occurs. Is that also true for an NJFET?

If 130mV is the absolute maximum signal level that can be accepted at the gate of the second stage, and it's likely that musical peaks will reach 150mV at the gate of the second stage, then the overload performance of this circuit is marginal at best.

But I don't know if the gate-source voltage defines the maximum signal level that the NJFET (common source) stage can accept. Anybody know the answer to that?

--
 
Rich tone = lot's of nonlinear distortion.

And? I'll reserve respect to the comments till after you try it.

The sound is considerably deeper and more revealing as well. Seriously, I may have overlooked LDR's had someone told me about the distortion. But now I know what they offer far outweighs the distortion figure (that's not necessarily a bad thing, just is)
 
'Deeper', 'revealing', these don't mean anything to me, I'm afraid. Please describe the technical advantages of using LDRs. Where are you placing them?

I don't have any 2SK170's. I do however have lots of shiny new BC550/560 BJTs and OPA2134/LME4870 opamps.

The input resistor, it's used on nearly all cheaper phono preamps. It's right before the first transistor/opamp. It's usually to provide some RF protection on signal.

If deeper and revealing don't mean anything, and only non-emperical modeling does, maybe you should try some electronic hobbies that don't involve music. I mean, news alert, you ever seen distortion pedal schematics? It's about what the ears get, not the graph. Modeling is a tool for indicators, not rules of preferablity.
 
So are you putting an LDR in place of the gate stopper resistor. Tell me, is it in darkness or in light?

It may be about what the ears get when you are talking about guitar effects units. But these are devices specifically designed to produce heavy amounts of distortion and colouration. What we are trying to do here is build a good phono preamp, not an effects unit :p !
 
I don't think it does. This topology is just a bad design that won't go away because people like it's apparent simplicity.

These preamps are JFET versions of passive EQ valve preamps that are popular at the moment. The reason why they are so terrible is because they don't have the hundreds of volts of supply voltage to give the necessary headroom that passive EQ preamps need to work properly.
 
I think I've figured out how to get THD approximations in LTSpice.

With 5mV input to the Boozhound preamp, the signal level at the output is 300mV. This is in line with common practice (and confirms that the gain is about 40dB). The signal on the gate of the 2nd stage 2SK170 is 12mV. The THD is 0.2%.

With 50mV input, the signal level at the output is 2.9V peak. The signal level at the 2nd stage gate is 113mV. THD is 1.95%.

With 100mV input, the signal level at the output is 5.4V peak. The signal level at the 2nd stage gate is 227mV. THD is 4.2%.

THD is pretty high, but would not be my biggest worry. Blocking distortion is. It's likely that big pops on a record would overload the 2nd stage, forcing it to draw gate current. This would charge the 0.1uF input cap, which would need to be discharged through the gate leak resistor (221k) over the next few cycles. You could decrease the value of the input cap, which would shorten the time needed to discharge it, but that doesn't cure the problem.

Just for fun, I ran these sims on a tube preamp circuit I've been working on. This preamp is meant for the Denon DL110 cartridge, which has about half the output of a typical MM cart. That means this tube preamp has twice the gain of the JFET preamp. So please pay attention to the signal levels at the 2nd stage and output, and don't pay too close attention to the input signals. I made them proportional to the first example (5mV input for the JFET preamp, 2.5mV input for the tube preamp).

For the tube preamp:

The grid bias on the second stage is something like -6V. Gobs and gobs of headroom.

With 2.5mV input, the signal level at the output is 387mV peak (about 46dB gain). The signal level at the 2nd stage grid (not gate -- remember, it's a tube) is 24.5mV. THD is 0.026% (almost nine times lower than the Boozhound, even with slightly higher signal levels).

With 25mV input, the signal level at the output is 3.87V peak. The signal level at the 2nd stage grid is 246mV. THD is 0.12%.

With 50mV input, the signal level at the output is 7.74V peak. The signal level at the 2nd stage grid is 489mV. THD is 0.23%.

Just for yucks, with 100mV input, the signal level at the output is 15.3V peak. The signal level at the 2nd stage grid is 984mV. THD is 0.46%.

Also, there's still no possibility of blocking distortion at this level, as there's no grid current drawn (remember that the grid bias is way up there at -6V).

Granted, this is a bit of an excessive design. But these sims seem to confirm that the NJFET preamp doesn't have much headroom. It has enough to get by, and is probably a nice project for the beginning DIYer. Personally, I'd look at Eli Duttman's RCA-redux phono preamp with Russian 6N2P tubes and a source follower output. But that would be a bigger project, for sure.

--
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Higher B+ and Rload and what Rs you may in the second stage is a way. With 34-35V B+ in typical FSP and stock RL, Rs it handles 16V pk-pk sine before visible clipping. 50 Ohm Zo. Had made 22V pk-pk tweaked builds also. 2V vertical scale in the picture, hitting the 15V you asked for cleanly. 0.02% @ 300mV measured on average for LMC builds.
 

Attachments

  • IMG_20150511_215837.jpg
    IMG_20150511_215837.jpg
    943.2 KB · Views: 392
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.