Grand Finale
This project has had a number of separate improvement efforts within it, of which the hum reduction effort was just one. Collectively, the improvements then amounted to:
1. Doing a complete make over of the RCA jack layout over that of the original design, providing 4 separate high level stereo inputs. Because the low level inputs were not going to be used on this unit, it freed up that jack area to help facilitate this goal (all the low level circuits were to be left in place and untouched as much as possible, but otherwise their input jacks and wiring removed, and power to these circuits removed as well). Cardas jacks and bare FR4 board made for a neat, clean revision of the jacks, with left and right inputs clearly marked by colored rings now around each jack. The FR-4 allowed the jacks to remain isolated from the chassis for lowest hum and noise.
2. The same treatment was given to the old Tape Output jacks, with FR-4 and Cardas jacks installed in this location as well. However, these jacks are now slightly reconfigured in conjunction with the old top chassis service switch. That switch now serves as a Preamp Output Switch, sending the preamp output signal to either the internal power amps, or out the old Tape Output jacks to serve as a Preamp Output feature. When the Preamp Output option is chosen, the power and phase inverter tubes may simply be removed, and the unit then operated as is a a full featured preamp, with most of the same features as that of the HF-85 preamp -- without any of the low level circuits of course. The switch prevents any external load connected to these jacks from affecting the internal signal path, or prevents the internal power amps from affecting the signal presented to the new Preamp Output jacks.
3. In finishing out the rear deck connections, the decision was also made to make for a more logical arrangement of the output connections as well.
Eico engineers were often too cute by half when it came to implementing the features of their equipment, many times placing themselves at odds with the majority of other manufacturers in this regard. Whereas others used channel designators of L&R or A&B, Eico used 1&2 for that purpose, reserving A&B for input options, where others used 1&2 for input options. The high level input jacks on the HF-81 are not laid out with any symmetry, with an odd number of input jacks, and one jack performing dual service serving one channel on one function, and the other channel on another function. During the days of AM-FM Stereo broadcasts and fledgling FM Stereo MPX, this was a useful feature, but most often today, it just comes off as confusing. To top it all off, when looking at the rear panel of the HF-81, the inputs were generally laid out with left channel inputs on your left, and right channel inputs on your right. However, this moment of clarity in the Eico think tank completely exploded when it came to output connections, with the left channel speaker connections and Tape Output jacks appearing on your right, and the right channel outputs appearing on your left.
To emphasize the level of utter chaos this general level of thinking achieved within the halls of Eico, the schematic of the HF-81 as published by Eico is only accurate if you understand that the upper phono preamp circuits are for the left channel, while the upper line, tone control, and power amp circuits are for the right channel!!! What were these guys thinking??
In trying to regain some sense of reality then, as mentioned previously, installation of the new Cardas RCA jacks immediately resolves the channel identification crisis for ALL of these jacks, while reassigning power amplifier channel identification now lets the left channel speaker terminal board be on your left, and the right channel terminal board be on your right. Consistency reigns for all connections now. What a concept!
4. Also as previously mentioned, the power switch is now located where the old Tape Speed switch used to reside -- minimizing wear and tear on the right channel treble control where the original switch was located (since it will no longer need to be rotated now to turn the unit on and off), and reducing hum by removing the AC power switch from the sensitive nearby tone control circuits.
5. All components have now been matched between the channels, if they weren't already. Performance has been verified to be virtually identical between the channels now, so that adding the Alps volume control allows for a very focused sound stage at all control settings now. Also, all the output tubes now idle at a very comfortable 10 watts dissipation each, with the tubes in each channel balanced for equal current draw by way of the new facilities to achieve that state.
6. All of the previous power supply and DC balance facilities have already been detailed to enhance performance, and reduce noise.
These efforts have really put this HF-81 in an elevated class of contenders now, with it not only very enjoyable to listen to, but to use and interact with as well. But still, that final layer of hum remained. What was annoying about it, was that it was no longer the low muffled tone of 60 Hz hum that covered up the other hum noise, as it had all been eradicated. What remained was that final low level, almost raspy, but undeniable hum that occurs from the magnetic energy running throughout the chassis, that could not be removed unless a way could be found to build all the circuits of both channels one top of each other at one point on the chassis. Since that's not possible, then the next best thing had to happen: The power transformer had to go.
Actually, all of the previous work is what made removing the transformer sooo successful, and, fairly easy to do from a stand point of execution.
Because all of the audio and power supply circuits had now been improved and/or reinstalled in a lowest possible noise fashion, removing the power transformer amounted to just that.
Normally, when a power supply is remoted, it not only includes the power transformer, but the rectifiers, filter circuits, and associated circuitry. In this case, literally, just the power transformer was removed, with all the rectification and filtering circuits remaining on board the HF-81. The results of remoting the transformer are really eerie as compared to the on board installation.
Normally, when you turn on a stock HF-81, the hum level goes through four phases rather quickly:
1. You immediately hear the magnetic coupling in both speakers, but particularly the one connected to the OPT closest to the power transformer.
2. As the output tubes begin to conduct, the magnetic coupling noise is diminished, replaced by 120 Hz hum from unbalanced output tubes.
3. As (or if) the output tubes start conducting more equally, then the 120 Hz hum is reduced, and the magnetic energy flowing in the chassis and amplified by the power amplifier circuits then begins to dominate and remain.
4. As the 12AU7 line stage tubes heat, then the magnetic hum is enhanced, and the unmistakable sound of 60 Hz AC heater hum comes in as an underlying foundation to the total hum noise.
With all the improvements already made and now the power transformer removed, when you turn on the amplifier, you get:
1. Initially.....nothing -- in either speaker. Dead silence.
2. As the output tubes begin to heat, you get......nothing -- whether the output tubes are balanced, or not. Still dead silence.
3. As the power amplifer circuits warm, you get......nothing.
4. As the 12AU7 tubes heat, you hear a soft hiss fade in as the tubes warm.
Turn the unit all the way up, advance the bass control to full, inputs connected or not, it doesn't matter. All that does is add to the timber of the hiss. It's doesn't bring out any otherwise muted hum noise. And all of this is with all the covers and skins off! There is literally ZERO hum in this unit now.
When the initial test of removing the power transformer produced such absolute results, the obvious move was made to properly remote the transformer. This portion of the project is still underway, but the pics provided will show the process and expected outcome:
1. This was a highly controlled test, with no expense spared to see if getting the damn thing out of the chassis would achieved the desired goal. It surpassed all expectations.
2. A suitable plate was then fashioned to cover where the transformer used to reside.
3. The power supply area was reworked slightly to accommodate the remote mounting of the transformer. The new front panel AC power switch on the HF-81 chassis is still active, as is the non-switched AC outlet on the rear apron. The umbilical is a 10 conductor 16 ga umbilical of nearly 6 feet in length, that is permanently attached to the HF-81 chassis. These efforts ensures no loss of voltage, particularly through the important heater leads.
4. On the new "power supply" chassis will reside the transformer, fuse, and CL-90 current limiter. Also there will be a standard IEC power cord connector, and 12 pin female Jones socket to accommodate the 12 pin male Jones plug on the umbilical. This arrangement ensures that no shock hazard exists, since AC power enters the power supply chassis first.
5. All parts have arrived except for the Jones plug for the umbilical. Therefore, the amplifier sits waiting and ready, with all skins back in place. The power supply chassis should hopefully be done tomorrow, with only the umbilical plug remaining.
The hum was very tolerable before the transformer was removed. But for those who want the last word in a black background, removing the power transformer is the only way to go. Just remember however, that it was but one of the required steps to finally eliminate all of it.
Dave