G9 and its frequency response ?? (impedance matching?)

[gyraf]

You may want to check your generator impedance - that is probably what will influence G9 frequency response the most.

The only place I've experienced lack of low end in G9/diy is at the highest gain settings, where the feedback capacitors somewhat limit response.

Jakob E.

 

[Jonte Knif]

Hi All,

I think people have been hearing and measuring (mostly) right because:
The problem is simple. LL5402 has way too low inductance for the output if you like your responses flat and bass without distortion. (obviously not everyone does and that is ok, I think) At 50Hz it has no load impedance around 1k (perhaps more, never measured one, lets take the data sheet as true). SRPP with ECC82 and the values shown in the schematics has Zout of 1,4k. That's were you get the bass roll off. The coupling cap has its part in the play in these kind of topologies, but in this case the trafo has bigger impact.

My simulation of the stage gives -6dB point at 50Hz when the load is 10k. Is this in line with real life measurements? Might not be quite exact.

A better alternative with ca 4 times the inductance (my measurement) would be LL1585, but it is different size. With 10u coupling cap it would show nice, extended, non-resonating bass resp. And the "poor SRPP" could produce undistorted bass at higher levels.

Kingston: tuu käymään joskus sun värkin kanssa, voidaan mitata kunnon vehkeillä jos kiinnostaa. (sorry for the encryption )


-Jonte

 

[rodabod]

If you are correct, then I'd maybe suggest adding series resistors before the output transformer and losing a bit of level in exchange for bass response.

 

[rodabod]

Actually, what I just said there might be nonsense. I'll finish my lunch and then actually look at the schematic!

 

[AudioJunkie]

This post jas me wondering, I now have an Intel iMac with VMware Fusion running XP better than any real Windows box I've used.  Does ayone know if I can ue my ULN -2 to test preamp responses with RMAA?  and if so, how? I have a Green that I built several years ago that I've always wanted to see its response and if Svart can get my G9 working I'd like to see it too plus I have a Focusrite Green I'd like to compare,

 

[Kingston]

My simulation of the stage gives -6dB point at 50Hz when the load is 10k. Is this in line with real life measurements?

Yes it is. I just woke up and I'm too lazy to check the RME HDSP analog I/O impedances, but yesterdays FFT spectrum analysis (from both trafo and DI input) show around 3-5dB bass cut at 40-50hz (that's as accurate as I can trust my tools), at all gains (ie. definitely not the feedback caps either). The bass roll off seems to start at around 200-300hz.

I will strip off the LL5402 and run some tests unbalanced to confirm mr. Knifs idea.

Kingston: tuu käymään joskus sun värkin kanssa, voidaan mitata kunnon vehkeillä jos kiinnostaa. (sorry for the encryption )

Kyllä kiinnostaa, kiitos vaan tarjouksesta. Mulla on tuossa muutama muuntaja vaihtoehto vielä kokeiltavana, mutta voisin jossain vaiheessa ottaa yhteyttä.

 

[rodabod]

Having looked at it a bit more, maybe someone could try just adding a simple pad to drop the impedance a little before the transformer. I can't see much else you could try when the transformer and valve type are fixed. I guess it must be worse with the OEP too.

 

[Kingston]

Having looked at it a bit more, maybe someone could try just adding a simple pad to drop the impedance a little before the transformer. I can't see much else you could try when the transformer and valve type are fixed. I guess it must be worse with the OEP too.

I'll try the pad. I actually have some of those 2:1 OEP's around so I'll run tests through it just for kicks. Guess now I'll have to post some data here as well, but at least we'll get this issue sorted.



The 10uF cap with a bigger output transformer certainly sounds like a desirable upgrade. I don't really care that they won't fit on the PCB as I've only done p2p tube projects this far anyway.

 

[Jonte Knif]

Fast checks:

ECC99 would be possible replacement for much better response and drive capability. With 10u coupling cap and around 12mA current it would perform fairly well. Pretty flat from 35 Hz.  Does not fit in 2 unit case But there are other tubes too to drop in...I think

5687 would fit in 2-unit, but needs mods in PCB and runs hot.

Both would lead to ca 700 Ohms Zout from the SRPP, which is quite optimal. Still lower and you start to have bass resonance due to coupling cap and trafo interaction. Any way, LL5402 is best for silicon carne in my opinion although the original SRPP can be pimped a bit. Don't take this too seriously, I don't like cap coupled tube outputs any way. I'm biased. Single-ended transformer coupling rules for many reasons. But that's another story.

-Jonte

 

[steppenwolf]

Hi!
I recently tracked a band and used the G9 a lot and for the first time I noticed some bass roll off, too. I compared it to a flat preamp and it does have less bass.
It would be really cool if there would be an easy fix for that "problem".
Best,
Stefan

 

[AudioJunkie]

This thread has me thinking, I now have an intel mac with VMware Fusion and it runs XP better than mot windows dedicated machines Ive used.  Does anoe kno if I can use my ULN-2 with RMAA? and if s how I canconnect it up to soe prees to test?
  I've got a DIY Green that I built a few years ago I'd like to see a freq response of and if Avart can get my G( build working, I'd love to compare it with my othr pres
Daniel

 

[tommypiper]

... I now have an Intel iMac with VMware Fusion running XP better than any real Windows box I've used.  Does ayone know if I can ue my ULN -2 to test preamp responses with RMAA?  

Welcome to the Mac!  Now you see...    It even runs Windows better...

Doesn't Metric Halo offer some testing and RMAA-type applications you can use with the ULN -2?

 

[gyraf]

Any way, LL5402 is best for silicon carne in my opinion although the original SRPP can be pimped a bit.

Is this just estimates, or have you actually measured the 5402's primary inductance when wired 2:1 like in the G9? With the 4u7 coupling capacitor I find it rather well-working, despite of what theory might have to say..

Jakob E.

 

[Kingston]

Any way, LL5402 is best for silicon carne in my opinion although the original SRPP can be pimped a bit.

Is this just estimates, or have you actually measured the 5402's primary inductance when wired 2:1 like in the G9? With the 4u7 coupling capacitor I find it rather well-working, despite of what theory might have to say..

Jakob E.

I've now made the measurements, and will post the data in the evening. The 5402 transformer is not suited for this application, and the OEP is only slightly better. I get "ruler flat" response from running the preamp unbalanced. the bass cut made by the transformer is also quite drastic (I consider -6dB at 50hz and error and a real problem, and not part of the sound), as you will soon see. And the cut is not affected by gain, or impedance of receiving equipment.

 

[Jonte Knif]

Is this just estimates, or have you actually measured the 5402's primary inductance when wired 2:1 like in the G9?

It was an estimate and a bad one! Shame on me. And shame on the data sheets too!

First, I thought that the no load impedance was given for series connection. It is not. And it seems that at least my unit of LL5402 had _much_ larger inductance than one could calculate from the data sheet. (ca. double)

Then, on the other hand the inductance of LL1585 is _not_ much higher. This really shocked me. It is better, but not much. Data sheet seems to give much too high value. Of course it would take more level, but that was not the point of the discussion.

I put one LL5402 into the same conditions as it would be in G9 (feeding from AP through 1,4k and 4,7u) and the freq resp was quite good. The 4,7u cap is right size to help to raise the bass with the resonance. I got 0,3dB drop at mid bass and very steep bass drop, almost flat at 40 Hz, but -2,5dB at 20 Hz. Bot high Z and low Z loads work well.

The price of getting straight response with the resonance is low Zin, and therefore the pre will (or should, I'm, getting paranoid) suffer from bass distortion at moderate levels.

So, everyone, do measure with very low levels to prevent the output stage from running out of current and getting flawed results because of that.

Now I am at least as confused as many other people here, because I get acceptable freq resp and for example Kingston not.

Oh dear. WHAT is going on?

-Jonte

 

[Kingston]

Right,

These tests are white noise sent from RME HDSP 9632 analog XLR output to G9, which is sent back to the RME XLR inputs. I did the same with sine sweeps, but the plot looks exactly the same. White noise was simply the faster way to test. It was sent with -15dB level (against 0dB digital reference, forgot to measure the exact voltage).

RME HDSP 9632 XLR I/O impedance's are fairly standard for most modern AD/DA: 10kOhm input and 50 ohm output.

Here I ran the preamp unbalanced (100k from the 4,7uF to ground). This picture also looks much like RME input directly to output, ie. flat, except the slight bumps and dips of the 4.7uF cap are visible. Unbalanced G9 sounds great by the way, if maybe even "too neutral", with one less trafo on the signal path. Hums a little, but it was good enough for this test.

Here, Lundahl LL5402. Both of the G9 channels match this curve, so it's not just a broken transformer.

Here's OEP A262-A2E, which surprisingly isn't that bad. The slight high cut corresponds to the sound, which I've found to be a little on the rough side.

Gain of G9 did not change these curves by the way. These measurements were made with moderate gain of each stage. Full gain looks the same. If I forgot to mention something crucial about the test data here, ask away and I'll try to fill in the blanks.


I did not try the pad before the Lundahl yet, and probably won't as there seem to be better solutions than loss of gain. There's already barely enough for ribbon mics.

I actually have some 5687 tubes around so I'll look into changing the SRPP a bit. My heater regulator is well heat-sinked so it'll (probably) handle the extra current just fine. Hopefully it won't need a major redesign, because this stuff is already somewhat out of my league.

In any case, these lundahls are almost too neutral for my needs(not counting the bass cut problem), so I'll have look at some alternatives from Jensen and Cinemag.

[edit]

additional note about impedance mismatches. I have an SSL bus compressor (stock standard NE553X [or equivalent] semi balanced I/O), which measures ruler flat in this kind of test. I ran the G9 through the SSL to RME. The bass cut was still there.

 

[steppenwolf]

@Kingston
If you should change something in the circuit please keep us informed about the changes!!
That is an issue to be sorted out!

 

[1954U1]

Kingston..
is the OEP's curve done with the OEP as transformer in?

I've done some tests with an AP, will post results very soon.

 

[Kingston]

Kingston..
is the OEP's curve done with the OEP as transformer in?

This is OEP as output transformer, and output of the G9 SRPP stage driving a transformer in general.

As far as I know, there are no known problems with any of the standard G9 input options.


Ok I should have added, in all those three examples, audio went through the LL1528 input transformer. I tested the DI as well, and response is flat there as well, as it should be.

 

[1954U1]

Ah, ok..
I have instead very similar results of the ones of mattnj & radiance..

transformer inputs [OEPs] -> big roll-off around 200HZ
DI inputs -> flat

cropping pics of the AP now, will post them with specs in 1 hour or so

Edit: turned out it was a bad connection on PCB. The roll-off now happens only with switch in "line" position