• Check out the Annual AK Secret Santa exchange Program, signup before November 23rd in the AK Charities & Christmas forum!

A Reference 400 Restoration/Modification

I'm not sure I'm following how you arrive at your Pd levels -- is the quiescent current range you are indicating just plate current as stated -- or is that cathode current? If it is in fact just plate current, then using a quiescent current average of 35.85 mA from your numbers and an average plate voltage of 417.5 volts from your numbers, produces an average Pd of 14.97 watts, or maybe not as cold as you were thinking.

If you're using the 400 volt plate/350 volt screen model in the RCA Tube Manual, that indicates a total quiescent current of 36 mA per tube -- but that's when the screen is at 350 volts, which it is not under quiescent conditions in the Fisher, with the screens in your unit running at 381 volts. The higher screen voltage the reason your total quiescent current should be up around 40 mA or more, so that when the screen voltage falls to 350 volts (or so) at elevated power output levels, then the quiescent current is more in keeping with the model presented.

Using a % of maximum Pd is certainly a safe way to go when setting the quiescent current, but works best when the specific % of maximum Pd is known for a given design. From a pure design standpoint, the optimum quiescent current is found from the transfer curve derived from the chosen load line generated when drawn over an appropriate family of plate curves for the tubes employed. But then that optimum current level must be modified as necessary as the specific installation so dictates -- that is, based on the characteristics of the power supply voltage supply points supporting operation of the tubes. If during any point of that design exercise the Pd (or any other element rating) becomes excessive at any point on the chosen load line, then a new set of operating conditions (and/or power supply characteristics) must be chosen. How much the power supply characteristics become a factor in that design exercise determines how applicable the installation of EFB would be .

In the 400 receiver, there are two points of opportunity, which both revolve around the screen grid voltage. As it drops with increasing power output, maximum power output capability is reduced, and because of the important relationship it has with the control grid, distortion is increased. Installing an EFB Screen Grid regulator will allow for the maximum possible power output that the power supply and chosen operating conditions is capable of allowing the tubes to produce, while installing an EFB Control Grid regulator will allow the tubes to produce the lowest possible distortion for the maximum power output a particular scenario will allow for. Using both together of course produces the best results, equal to that produced were the supply points fully regulated to specific set points from the same raw power supply.

In early 400 receivers, EFB will generate about 10 more watts of power output, and lower distortion to at least one half to one third of the original design. Whether that improvement in performance is of significance to your needs then is up to you.

Dave
 
dcgillespie said:
I'm not sure I'm following how you arrive at your Pd levels -- is the quiescent current range you are indicating just plate current as stated -- or is that cathode current? If it is in fact just plate current, then using a quiescent current average of 35.85 mA from your numbers and an average plate voltage of 417.5 volts from your numbers, produces an average Pd of 14.97 watts, or maybe not as cold as you were thinking.

If you're using the 400 volt plate/350 volt screen model in the RCA Tube Manual, that indicates a total quiescent current of 36 mA per tube -- but that's when the screen is at 350 volts, which it is not under quiescent conditions in the Fisher, with the screens in your unit running at 381 volts. The higher screen voltage the reason your total quiescent current should be up around 40 mA or more, so that when the screen voltage falls to 350 volts (or so) at elevated power output levels, then the quiescent current is more in keeping with the model presented.

Using a % of maximum Pd is certainly a safe way to go when setting the quiescent current, but works best when the specific % of maximum Pd is known for a given design. From a pure design standpoint, the optimum quiescent current is found from the transfer curve derived from the chosen load line generated when drawn over an appropriate family of plate curves for the tubes employed. But then that optimum current level must be modified as necessary as the specific installation so dictates -- that is, based on the characteristics of the power supply voltage supply points supporting operation of the tubes. If during any point of that design exercise the Pd (or any other element rating) becomes excessive at any point on the chosen load line, then a new set of operating conditions (and/or power supply characteristics) must be chosen. How much the power supply characteristics become a factor in that design exercise determines how applicable the installation of EFB would be .

In the 400 receiver, there are two points of opportunity, which both revolve around the screen grid voltage. As it drops with increasing power output, maximum power output capability is reduced, and because of the important relationship it has with the control grid, distortion is increased. Installing an EFB Screen Grid regulator will allow for the maximum possible power output that the power supply and chosen operating conditions is capable of allowing the tubes to produce, while installing an EFB Control Grid regulator will allow the tubes to produce the lowest possible distortion for the maximum power output a particular scenario will allow for. Using both together of course produces the best results, equal to that produced were the supply points fully regulated to specific set points from the same raw power supply.

In early 400 receivers, EFB will generate about 10 more watts of power output, and lower distortion to at least one half to one third of the original design. Whether that improvement in performance is of significance to your needs then is up to you.

Dave
Click to expand...
 
Without getting too long winded in my explanation let me just say the the data I sent came from two different sets of data from two different tests which understandably would cause confusion for anyone who did the math. My apologies for that.

I am in agreement with the calculation in your reply, but it didn’t explain why, based on the RCA tube data, my perception was that the tubes might be running cold, that is, until I noticed something that I may have never realized. There is a note at the top of the characteristic chart that says “typical operation (values are for two tubes)”. Does this mean that the total “zero signal plate current” under the 400 plate volt/350 screen volt column of 64ma needs to be reduced by half for a single tube? If that is correct it would mean that the note could only apply to current levels, not voltage levels, since we all understand that the nominal plate voltages, for instance, would be theoretically identical. And if so, that would be why I was concerned about tubes running cold.

I also now understand that the maximums on the characteristic chart are for the tubes when “in the box” since it couldn’t possibly take into consideration any circuit design elements in which the tubes might be used. So once again your explanation, although at least a small portion of which might be a little above my technical expertise, goes a long way toward clarifying that for me.

As far as whAt my needs are regarding whether to move forward with the modification, sometimes I wonder just that. I don’t think I have any real “needs”. I decided to use any restoration and/or modifications as an exercise in reinforcing and expanding my knowledge from when I moved away from my much earlier electronics training in high school. I never had any expectation of what would be done with the 400 once it was completed. Most likely I will keep it but I am no audiophile and I do not listen to music like an audiophile would, so I could conceivably leave the unit the way it is and be perfectly happy with it. But then I wouldn’t be learning anything like I am right now.

I think that I WANT to consider the EFB upgrade so if it wouldn’t be too much too ask to pencil sketch the circuit as offered (for the earlier version I) at your convenient, that would be most appreciated. If the differences are minor, perhaps some component value changes, it might be easier to mark up the original schematic for me to follow. Schematics have ever been difficult for me to follow. Thanks again for all you do.

Bob
 
Hi Bob -- Yep, the current draw levels given in basically any tube manual depicting push-pull operation is for both tubes, so you would divide by 2 to get the single tube levels for either tube in the pair.

I should be able to post something in the next day or so.

Dave
 
After having 2 units with EFB I'm building another amplifier that will have it too.
One of those 2 units is Fisher 400. Both are just barely warm above tube area at my listening levels and they sound great. Not much hotter then big SS beasts!
EFB is a great approach to convert energy into sound, not into heat.
Thanks Dave!
 
Hi grind. I’m curious if either of the fisher 400s are of the earlier variety with the non -AX opt part numbers? Dave has offered to send me a sketch of the EFB mod for my early version receiver which I am sure will differ somewhat from the design he has published some time ago for the later version. I am awaiting that info. In the meanwhile if you have done a mod of an early unit I would love to hear feedback on the outcome.

Thanks, Bob
 
Can anyone recommend what would be a respectable speaker to go along with the fisher 400? Not looking to mortgage the house to get top shelf, just something that respects the quality of the receiver, possibly a used set from the period if that is reasonable.
 
Bob Vecchi said:
Can anyone recommend what would be a respectable speaker to go along with the fisher 400? Not looking to mortgage the house to get top shelf, just something that respects the quality of the receiver, possibly a used set from the period if that is reasonable.
Click to expand...

My 400 is late model version.
Most vintage Klipsch speakers will sound good with Fisher. Cornwalls, Forte, Chorus and most likely others. Most of horn loaded speakers will sound good with tube amplifier due to their higher efficiency and lower damping factor requirements.
 
Wow, it seems like I may be the only one on the planet who has an early version. Does that make it more valuable, lol? Thanks for the help with speakers. Never hear of Cornwall’s until I came on here, nor Forte or Chorus. Klipsch I am familiar with. I w
Had my eye on some JBLs but they have cone tweeters. I may get them because the price seems right. But I will continue my search. Thanks again
 
Here you go, which would be the latest implementation of EFB™ for an early model 400 receiver. The two incoming arrows on the top left of the schematic connect to the 25 VAC leads that connect to the bridge rectifier. The bridge and filter caps and their connections all remain as originally designed. These two leads now represent additional connections to the 25 volt source. The four arrows coming off the four pots on the right side of the schematic go to the individual grid return resistors as usual.

Dave

Early 400 EFB.jpg
 
Larry, I am curious to know the how the results of the EFB/ IBAM mods in the two units compare, and your opinion on how the mod benefitted the health of the early version receiver specifically.
 
Both have the IBBA (Dave's Bias/Balance board) vs. IBAM. Neither has the EFB..........yet. Still de-cluttering after moving last year. Bias is very stable with the IBBA board, and the pairs bal+/- 0.01v max. btwn tubes in each pair. With them at 75% of max dissipation, they are a lot cooler vs. a stock unit. Less strain on the circuit and tubes overall. Granted this is all by feel and ear. They aren't room heaters anymore. Output's are now very Warm(20 Seconds very warm), vs. HOT.(Lose fingerprint hot!) The early one sounds a bit smoother than the later one. Both are running Sylvania tubes in the RF/IF sections, Sovtek 12AX7LPS in Phono 12ax7 Fisher branded tele's everywhere else, and EH 7868's. I've run Tung-Sol 12ax7 Russians in both @ the P.I. but went back to the Tele's (too bright).
 
Dave, I reviewed the schematic you sent, and being curious, I compared it to the schematic for the later version. Of course I was expecting some differences (otherwise I wouldn’t have needed it,right). Anyways, in preparing the BOM I noticed a diode across the .1 cap from the base of the first ksa to ground. That seems odd to me especially since the other schematic which is basically the same has no such diode. Is this correct? Also the later version drawing has 3 notes that I assume do apply to this drawing as well? Is this correct? It occurs to me that it’s been a long time since this mod was designed and that you may have inadvertently left the notes off. Or you may have assumed that I would be smart enough to figure it out for myself (this type of thinking will get you in trouble). I’d rather get it right before moving forward.
 
The diode was added to help protect the B-E junction of the inverter transistor at shutdown in certain applications, but now I just added it in regardless. In some applications, at shut down the negative supply voltage can disappear quite quickly, while the B+ voltage may linger. In that instance, the base of the inverter transistor will be driven positive, which is not exactly something it enjoys being a PNP transistor. The diode is back biased during normal operation, and so appears invisible to the circuit. But at shut down, its there just in case to protect the base junction should it need to. Because of the rather unique full wave voltage doubler I developed for use with EFB when it is installed in Fisher receivers, it's presence is really optional as there's never been a failure of the inverter transistor in that application. But, owing to my effort to always make a design as bulletproof as possible, I just include the diode automatically now.

There are other differences as well, some of which are improvements (like adding the gate stopper and upping the value of the gate cap), but others specifically because of the differences in the output stages of the two versions. Regartding the three notes:

1. Note #1 only applies to later version installations because the screen grids in that version should be operated at a much lower voltage than the original design provides for. As such, in the later version, the EFB Screen Grid regulator is only powering the output tube screen grids -- this being another bulletproof effort by minimizing dissipation within the MOSFET. In early versions however, the screen grid quiescent voltage as provided in the original design is entirely proper, so in that case the value of R105 (originally a 1.2K 7 watt resistor) is increased to 2.5K at 5 watts (shown on the schematic provided), and the EFB Screen Grid regulator is effectively connected in parallel with it. With this configuration then, the resistor and regulator together are effectively powering the entire set (other than the output tube plates), with the resistor acting to eliminate needless dissipation in the mosfet under quiescent conditions that it doesn't need to dissipate during that time.

2. Note #2 obviously applies to all versions.

3. Note #3 has been discussed in the text previously as being ideally 38 mA in early version receivers.

Also of note, later version EFB installations do not specify Screen Stability resistors, because in that application (i.e., specifically when EFB is installed), they really are not required even though they typically get included as a matter of recourse anyway. In early version receivers however, the installation of such resistors is mandatory, whether EFB is installed, or not.

I did not go into the usual detail I normally do, as this was not a formal presentation I originated but rather, simply a response to a request for how the published EFB circuit for late version 400s is modified to operate in early version 400 receivers. To that extent, I have never formally presented an EFB installation for any of the Fisher's receivers but late version 400 units, as the need there was so great. The EFB information for other receivers that's out there (at least 500C/800C) was generated like this one was -- as a individual request (by PM as I recall) I performed that was later worked up into a formal schematic and presented by the individual requesting the information. In those occasions, I have to assume that those wanting to take on such an advanced modification have a good working understanding of circuit action, and can make the necessary application adjustments on their own, or with minimal guidance, or with the help from others. I hardly mean this to be taken as any form of complaining or chastising, but merely as an explanation as to why the level of detail is as it is. I always consider myself privileged to help anyone who is willing to take on any of the circuit modifications I have presented over the years.

Dave
 
Thanks for the clarification. No offense was taken. On the contrary I am grateful that you took the time to do it. I ordered everything and can’t wait to get started.
 
Nice to see this all concisely summarized in one thread. I for one can vouch for it taking a lot of time to read and parse together all the information previously presented in multiple threads.

dcgillespie said:
o in that case the value of R105 (originally a 1.2K 7 watt resistor) is increased to 2.5K at 5 watts (shown on the schematic provided), and the EFB Screen Grid regulator is effectively connected in parallel with it. With this configuration then, the resistor and regulator together are effectively powering the entire set (other than the output tube plates), with the resistor acting to eliminate needless dissipation in the mosfet under quiescent conditions that it doesn't need to dissipate during that time.
Click to expand...
Dave, I'm curious how you decide how much voltage drop the mosfet should see under max power conditions? In my EFB installations (none of them have the parallel current shunting resistor), I put about 20V across the mosfet at max power, and thus (try) to size the series voltage drop resistor sitting right in front of the mosfet accordingly so the mosfet always has min 20 volts across it. Of course the mosfet has more volts across it at idle since the screens conduct less current at idle. Is 20V about the right amount? Or how do you decide?
 
Kevin -- Since the specified mosfet is so overrated for the typical applications that EFB is involved with (particularly the non-plastic versions), and so electrically protected within the EFB model, I primarily concentrate on mosfet dissipation, since that is about the only thing that could take it out. With a little trial and error, it's not too difficult to determine a dropping resistor value that essentially maintains a constant dissipation from the mosfet, from quiescent conditions to full power output in both channels as the case may be. I am typically able to keep electrical dissipation to under 3 watts which, coupled with the safeguards built into the circuit, pretty much have made for a no-fail condition for the device, which governs just about everything and anything I do involving SS electronics.

I hope this helps!

Dave
 
You must log in or register to reply here.
Back
Top Bottom