Lewm exactly! Well said abs better than I could put into words.
I used VCap ODAMs in that position. Easy surgery.
I used VCap ODAMs in that position. Easy surgery.
JB, On the subject of changing capacitors, the basis for choosing the value of the output coupling capacitors is to provide for the best possible extreme low frequency response, first of all. (f = 1/2*pi*R*C, where R is the net impedance seen by the capacitor in Ohms, and C is the capacitance in Farads). That calculation determines the minimum value of the coupling capacitor, depending upon where you want to cut-off the LF response. The next consideration is the overall SQ of the capacitor that you choose. In this case, you say you switched from the OEM 30uF capacitor to 3.3uF and then to 12uF. 30uF is going to be way overkill using any imaginable amplifier, in achieving a high pass filter that goes down below ~2Hz, which is more than adequate. But lower values of capacitance can also give you a 2Hz cut-off frequency or near enough to it in the real world. That depends upon the parallel sum of the impedances that the capacitor works into. Like I said above, nearly any amplifier you might own (solid state or tube) will have at least a 10K ohm input impedance. Most will be higher. 10uF capacitance is sufficient yield a ~2 Hz high pass filter into 10K ohms. (Insert the values into the equation above, to see what I mean.) If you are driving a tube amplifier with a more typical 100K ohm input impedance, then even a 1uF capacitor will do the job. The next question is SQ. Yes, it is possible that your 12uF capacitor just happens to sound better than your 3.3uF capacitor; you haven't stated the types or brands that you chose. In that case, the 12uF will sound better simply for the fact it IS better, not because a higher value is necessarily doing you any particular good for low bass response. In the case of your Chinook and my Steelhead, the OEM coupling capacitors are 30uF in value and the brand is either REL or MIT (made by REL) metallized film type. REL make some wonderful capacitors, but this particular type is not among them, in my opinion. So I made my changes not only because there is no point in using such a large value capacitor in this application, but in addition because the particular 30uF capacitor they chose is not optimal with respect to SQ, in my own opinion. Manley were concerned with cable capacitance, when they chose to use such a high value of coupling capacitor and to insert the 47 ohm resistor in series with the output. Their choice enables the units to drive very long runs of cable with high capacitance, without messing up the frequency response, but I just viewed their choice in that case as overkill. I would never use such long and capacitative cables in the first place. Many manufacturers make choices like this simply because they count on the end user abusing their product, e.g., by asking the preamp to drive 40 feet of high capacitance cable. Then the end user comes away critical of the product when it fails to perform well under such a ridiculous stress. |
I also replaced 3 of the 4 ceramic cast resistors on the power supply side for Mills 10w military spec. I removed the 3.3uF and put in a 12uF. I think the 12uF sounds better, but I'm not sure why. Maybe Manley was on to something with the 30uF--a little bigger than normal or calculated values sounds better. I used VCap ODAMs and had to wrap them in silicone tape to be sure the bodies do not conduct. Finally, one of my Sparkos labs discrete op amps went bad last year so I'm back to the Burr Browns. I think the cap upgrade sounds better than the Sparkos. I'm going to email Sparkos and ask for a refund or replacement as there's no reason it should die in a year. |
I took Lewm's advice, but I also did a bit of math and consulted some smart dudes on DIY forums. I went from 30, down to 3.3uF, as it did appear that cap was way overkill. Still breaking things in (and I only dry fitted the 3.3uFs). Things are punchy and pretty much within the sonic signature. Very interesting. I'll need time to assess. |
etnier, My unit is a Steelhead. I am not sure that it is configured exactly like a Chinook, so you would need to confirm that first. I was also unable to obtain a schematic of the Steelhead. Even my professional tech who runs a thriving repair business was unable to obtain it. My insight comes from a verbal description of the circuit provided by Evanna Manley in the context of a review of the Steelhead published years ago in 6 Moons. According to EM (and confirmed by my investigating the circuit with an ohmmeter), the 6922 is the only tube involved in the phono equalization and gain circuitry. The two 5687s are used as cathode followers (CF), as follows: Output from the phono section goes through the first CF, which is actually a complex type called a "White Cathode Follower", requiring both halves of one 5687. This drastically lowers output impedance and drives the volume control via a 47 ohm resistor in series with a 30uF capacitor in the original design, if you use the Steelhead as a full-function preamplifier, as I do. (If you use it only as a phono stage, bypassing the volume control and other switches, then the 47-ohm resistor/30uF cap would be driving the input of your linestage.) According to my meter, the volume control has a 5K ohm input resistance, which is very low compared to most. However, 30uF is very high value for a coupling cap, even one that has to drive a 5K load. And it's way higher than needed to drive any reasonable linestage. Manley used a (mediocre in my opinion) metallized polypropylene type here. By my calculation, all you need is 12uF into 5K, to get a hi-pass filter at 2Hz. The closest to that value and best I had in my stash is a 10uF/200V MIT polypropylene film and tin foil type, far superior to the original cap in SQ. (Last time I looked, Michael Percy sells the MIT caps.) So I used that in place of the 30uF cap. (The difference between the calculated 12uF and 10uF is academic at best.) The 30uF cap and the 10uF MIT cap are about exactly the same in size, so no problem with fitting it in. I also bypassed the 47-ohm resistor; this serves no needed function unless you are driving VERY high capacitance cables to a linestage, and I don't. It's totally superfluous for driving the volume control. The volume control drives another WCF using both halves of the second 5687, which drives the input of an amplifier, in the case where you use the Steelhead as a preamplifier. Most amplifiers have an input impedance north of 10K ohms, if solid state, or north of 50K ohms, if tube type. The Steelhead has another 47-ohm resistor in series with another 30uF capacitor at its output here. The 30uF capacitor is once again way overkill for driving even 10K ohms. I deleted the 47-ohm resistor and replaced the 30uF capacitor with a 4uF/250V polystyrene and tin foil coupling capacitor. (I had these in my stash; you cannot buy them any more, but you would use the best possible 2 to 4uF capacitor here, if you are driving anything above 20K ohms.) The difference is "startling". |
I didn’t have any reservations about changing the output coupling capacitors in my steelhead, because it is several years old and I am not the original owner. Thus I assume that I have no warranty in place. However, Manley seem to be very in tune with tweakers. I wonder whether they would sanction capacitor upgrades if done by a qualified tech or at their factory. you can’t go wrong spending 69 bucks on a PC. If you don’t like it, there’s no harm. |
Apologies to slaw, but I have come to think that SR is throwing sand in the public eye with $150 fuses and many (but not all) of its other tweaky products. While I also question the bang for buck value of very expensive power cords, at least there you can objectively identify desirable characteristics and aim to achieve them. I own a Steelhead. I’ve thought about upgrading the load resistors to nude vishay, but my prior experience with that sort of tweak on other phono stages suggests to me that there is little to be gained. I could be convinced otherwise. Like I may have written above, the major kludge in the Chinook if the output stage is like that of the Steelhead is the choice of output coupling capacitor. The value is way higher than necessary for good bass response and the quality is mediocre at best. Upgrade those and you will be amazed. |
@soundermn , The cost of that fuse was a major hurdle I personally had to overcome. In the end, it was a great move. The Cable Company has a great sale every now and then...buy the outlet, get the fuse free. Myself...I’m waiting on it one more time. We have to overcome, as humans, the effect of some products having a great effect/while being small/simple verses the cost. Once you experience it....you’ll understand. |
If you are reading this in the future, this is an interesting read about op amps. I’m not expressing an opinion, only sharing what I found. http://nwavguy.blogspot.com/2011/08/op-amps-myths-facts.html I bought a new new set of tubes to roll, and the signal cable power cable. After I get those going I will report back... and I’ll decide about the op amps. |
@johnss I think you mean the duals. Those suckers are $80 each?! So that’s almost $200 after tax and shipping. Is the difference that great? For that much I could buy a set of NOS tubes, and they are in the signal path, so can have a huge impact. The stock opamps are about 30 cents. So maybe something that costs around $10 to $15 would be a nice improvement without going overboard. What do you think? |
@slaw I agree that I don’t want to do any soldering or anything that would void warranty. I can see how they could help, but I like warranty! At the same time, I’m not convinced by a $150 fuse. It could work, but hard for me to conceive. I’d try it for $30 but not $150. @mattmiller 👍 yep. I ordered a signal cable digital PC. Should be here in a few days. @johnss this sounds interesting. And, I think it’s more likely to make a difference using the MC setting, as I am. Since it drives the higher gain input, I can see your point and understand why it could help. I think I will try that next. Did you get the dual or single op amps? |
the Chinook sounds very clean in stock form. but when you insert the discrete op amps, the signal is really cleaned up. it makes the stock parts sound out of focus. Correct, they are not in the signal path, but do control the voltage going to the JFETs in the first stage. The other big improvement on the Chinook is to replace all the MC and MM load resistors (located on the back PCB bolted to the rear panel) with Vishay naked or Vishay VSR bulk foil resistors. |
Okay I use a Signal Cable "Magic" Power cord with my Manley Labs Chinook and it makes a big difference. What I found was a much tighter overall sound if that makes sense, everything improved. Also, putting some brass weights or something to dampen the metal screen casing from vibrating that is a key tweak, imo. Keep in mind that the Chinook is not a heavy piece of gear and the Signal Cable IS, the connection and cable are massive be ready to support it in someway also order 2 feet more than you need this stuff dosent bend easily! Matt M |
Although I haven’t tried a cord with the Chinook, IMO, your best bet would be to install a Synergistic Blue fuse. This will definitely give you bang for the buck compared to any PC out there. Then get the Blue outlet! Still, if you're looking for a heck-of-a-bargain in a power cord, try the latest Signal Cable silver alloy. At $169 shipped (mine had a 20amp IEC on it), it may be the best deal going. It replaced my long time Shunyata Python Alpha Helix 20 amp feeding my Hydra. Increase in transparency with no negatives at all. I think the Python retailed for $1k. |
I have the Chinook and have tried several aftermarket cords. Can't say I can really tell much of a difference. Power cords seem to do a lot more with SS gear than with tubes. For the Chinook, if you want an easy upgrade, swap out the TL082 op amps at the rear of the unit and replace with Sparkos Labs discrete modules. The remove several curtains of grunge over what the TI parts do... and best off all, no soldering, simply unplug the stock units and insert Sparkos modules. well work it |