Improvments for the HF-81 phono preamp section

[CZ4A]

I am in the process of recapping my Eico HF-81. I have read that the built-in phono preamp section leaves a lot to be desired. For now, I am just doing a 1-for-1 replacement of the electrolytics and resistors with film caps and Kiwame resistors. Besides component upgrades, what else has worked to improve the phono preamp section on this amp?

 

[sKiZo]

The only major mod I've seen to the stage is to relocate the EQ section under the chassis to minimize wiring runs, and to improve the shielding.

http://home.earthlink.net/~eico_hf81/eico_hf81_phono.htm

I'd think you'd be able to accomplish most of the stated benefits by just adding a shield to the existing configuration? You should be able tweak most of the little bits on the gang switch and add a full height metal wall for complete isolation.

 

[Selmerdave]

I don't know how similar it is to the phono section in the HF-85 but years ago I did mods that were detailed in the Yahoo Eico group. These included moving the RIAA components to the phono jacks as mentioned above, and I was happy with the improvement. Not sure if there's something there similar for the HF-81.

 

[dcgillespie]

The phono preamp circuit in the HF-81 and the HF-85 are identical. They are an active (NFB) design, but unique in that the feedback is only applied around the second stage.

Dave

 

[ferninando]

have you tried the eico forum?

 

[ferninando]

I moved all the HF85 phono RIAA parts right onto the phono tubes on perf bds.
naver have used it but it should work. everythimg else i rebuilt did.
see the 2 small perf bds with some blu parts. thats the RIAA EQ.

 

[CZ4A]

Thanks for the suggestions. Right now I just want to get the HF-81 up and running so I'm not messing with the RIAA section yet, but I will definitely keep these ideas in mind if/when I'm ready to tackle it. I have a U-Turn Pluto pre-amp anyways. That will make for a good comparison point before and after RIAA mods, I think.

 

[drbubble001]

If one were to transplant phono RIAA eq (90k/.21uf+750k/.0027uf) from selector switch to phono pre, would it be inserted form points directly after C1 (.25uf) to juncture directly after C7 (.025uf) - Eico HF-81 schematic? Thanks. Somewhat new at this, excuse possible inaccurate terminology.

 

[tubeactive]

Welcome to AK @drbubble001 , The RIAA feedback network values you wrote must need an edit, because they do not represent an RIAA network.
Having restored and retrofitted my HF-81 and HF-85 phono stages, I recommend this network as an upgrade to the original values, which are close to ideal: 2700pf//2.2Meg. Ohms & 820pf//91K Ohms. This has been objectively verified to be very accurate.

It is simple to measure R and C values with VOM and LCR meters, respectively, in order to match the parts per channel. However, retaining these values after soldering requires extensive heat sinking methods. I use very small, copper, flat bladed alligator clamps, attached to the part to be soldered, as well as attaching them to any adjacent, touching parts, to keep the precision EQ values intact.

I believe your partial schematic is from an Eico manual. I think your chosen insertion points are correct, but your labeling of start/send and return are flipped. The feedback "starts" by the .025uf cap and returns to the V2 grid. I will dig out my factory scheme and check again...

For Eico lurkers, the phono EQ circuitry on HF-85 and HF-81 Eicos is identical. The HF-81 has one stereo volume control and one balance control, with no Loudness control. The HF-85 preamp uses two volume controls, one per channel, concentrically mounted, with taps for a switchable Loudness control.

Unfortunately, the HF-85 Loudness circuitry was published in error and does not sound right in stock form. However, as a saving grace for me, these loudness taps are a great placement for a scratch filter to use with noisy discs and sources. Since the Eico volume control is directly after the phono stage, before any other inputs, it was logical to implement a high-cut, scratch filter with little to no phase shift of the signal. Usually, scratch filters, and tone controls are placed after further gain stages, causing possible phase shifts at certain frequencies, thereby adding musical coloration changes...

AudioKarma does have a superb "Eico Place" thread section, plus the Turntables section is reserved for "All Things Analog" including preamplifiers and phono stages....You might want to check these other AK sections and use the easy Search function to gather more preamp info...

 

[dcgillespie]

Tube is exactly correct about the EQ values and Send and Insertion points for the RIAA EQ network. As measured, the accuracy of the network is in fact pretty good, although it rolls off fairly quickly below about 36 Hz -- where little recorded information exists on most vinyl recordings anyway.

If I may however, rather than being an error, the loudness circuits of (most often cited) the HF-85 and ST-70 actually work very closely to the "intended" purpose of that function. I'm not saying that everybody or anybody should like it -- just that it is not an error in design. It was Eico's intended purpose to make their models operate as they do with regards to this function.

The concept of loudness control is to alter the response of an amplifier at low volume levels -- read that again: LOW volume levels, meaning background level sound levels or LESS. Strickly speaking then, it is not the Eico designs that are in error, but in fact everybody else's. Most/all other amplifiers do not cause such a drastic drop in volume when the loudness function is activated, nor add in such extreme levels of LF boost. But that is exactly what it is supposed to do: The significant drop in volume then allows much more accurate control of the sound level by the volume control when operating at low sound levels, and the bass is appropriately boosted to compensate for the ear's reduced response to LF information at low listening levels.

Nobody else however followed that lead. Instead, their loudness controls do not drop the sound level at all, so that the LF boost is added in well up into foreground to loud listening levels and beyond. Some even add in a slight HF boost as well. Adjustment of very low sound levels becomes more difficult however, because the volume control then becomes very touchy and often mis-tracks at such low settings. But with so many other manufacturers designing their loudness function in this manner then, the way they operate became the norm and even expected norm. Against this, the Eico loudness function seems to be in error -- an engineering error, a schematic error, a kit error, or builder error -- but none of this is true. It was Eico's effort to in fact mimic the true intention of what the loudness function was supposed to do. The problem is, their message never got out. It wasn't something they promoted in advertising, nor discussed in any of the magazine articles their engineers wrote about their products. And with all of the competitor's pieces operating uniformly the same way, but distinctly very different from how the Eico design does, then who looses in that scenario? Eico does, with someone suggesting long ago that there was a schematic error, and so the story went, and grows.

I have no dog in this hunt if anybody and everybody changes an Eico's loudness function so that it more closely mimics the customary operation we normally expect from this function -- I would no doubt change it as well, largely because nobody of our interest level listens at such low listening levels (we're too interested!) except maybe late at night. My only goal is to point out that the Eico loudness design was not an error in design or schematic value. It was Eico's intention for it to operate that way -- and to heck with how all the other manufacturers did it -- whether the listening public liked it that way or not. As history clearly shows however, they clearly lost that battle long ago.

Dave

 

[tubeactive]

Thanks @dcgillespie ! With the Loudness switch unhelpful in my system, and I rarely like loudness circuits, I adapted to my phono playback needs. When I retro-restoed my Eico, I used the function switch wiring and the Tape EQ switch to my advantage, implementing three high gain phono inputs. In my retro-resto, the Tape EQ slide switch enabled two different phono playback curves, the Col. LP/NAB and AES phono EQ curves, needed for many mono discs...

Concerning bass reduction below 36Hz on the HF-85/HF81 units, when and where was that encountered ? Using parts matched to 2% or better, the response thru my entire HF-85, including line level gain and tone stages was rather exemplary, objectively tested by a more senior sound engineer.

When I tweaked the Eico phono networks, I only adjusted the Treble Rolloff values. Of course, with active feedback EQ, the two R-C networks in series "combine and interact" to derive the Turnover, but my RIAA EQ was quite accurate, with no bass rolloff before 20Hz. Eico did design in some low pass filters in line stages, though, which could cause some rolloff after 14KHz, but I tweaked those networks for 19KHz...I enjoy the HF-85...

 

[dcgillespie]

Tube -- It is quite possible your unit produced better measured results than the unit I measured did. I have no doubt your engineer friend measured the circuit properly, as did I, using a Reverse RIAA Network of known accuracy (within 0.10 db across the full audio spectrum when driven from a 600Ω source) -- although I measured just the phono preamp circuit itself so that the following tone stages could not introduce any error of their own. My test was conducted with the phono position selected (so that the volume control loaded the phono section normally), with the output of the phono preamp directly driving a buffer amplifier with an input impedance of 22 megohms, which then drove the test equipment so as not to unduly load the output of the preamp down by the input impedance of the test equipment.

The real issue is the fact the RIAA curve has a dynamic range of 39 db, but the stage that the EQ network wraps around has at best a gain of about 36 db using a perfect tube of bogey specifications. So the ability to reproduce the curve properly is compromised to begin with. Then if any element of the circuit is operating less than optimally or perfectly, the low end response will suffer, more than what it already inherently suffers due to the lack of open loop gain. Given that scenario then, measurement discrepancies from one unit to the next would be very possible, if not in fact very likely.

Dave

 

[drbubble001]

Welcome to AK @drbubble001 , The RIAA feedback network values you wrote must need an edit, because they do not represent an RIAA network.
Having restored and retrofitted my HF-81 and HF-85 phono stages, I recommend this network as an upgrade to the original values, which are close to ideal: 2700pf//2.2Meg. Ohms & 820pf//91K Ohms. This has been objectively verified to be very accurate.

It is simple to measure R and C values with VOM and LCR meters, respectively, in order to match the parts per channel. However, retaining these values after soldering requires extensive heat sinking methods. I use very small, copper, flat bladed alligator clamps, attached to the part to be soldered, as well as attaching them to any adjacent, touching parts, to keep the precision EQ values intact.

I believe your partial schematic is from an Eico manual. I think your chosen insertion points are correct, but your labeling of start/send and return are flipped. The feedback "starts" by the .025uf cap and returns to the V2 grid. I will dig out my factory scheme and check again...

For Eico lurkers, the phono EQ circuitry on HF-85 and HF-81 Eicos is identical. The HF-81 has one stereo volume control and one balance control, with no Loudness control. The HF-85 preamp uses two volume controls, one per channel, concentrically mounted, with taps for a switchable Loudness control.

Unfortunately, the HF-85 Loudness circuitry was published in error and does not sound right in stock form. However, as a saving grace for me, these loudness taps are a great placement for a scratch filter to use with noisy discs and sources. Since the Eico volume control is directly after the phono stage, before any other inputs, it was logical to implement a high-cut, scratch filter with little to no phase shift of the signal. Usually, scratch filters, and tone controls are placed after further gain stages, causing possible phase shifts at certain frequencies, thereby adding musical coloration changes...

AudioKarma does have a superb "Eico Place" thread section, plus the Turntables section is reserved for "All Things Analog" including preamplifiers and phono stages....You might want to check these other AK sections and use the easy Search function to gather more preamp info...

Welcome to AK @drbubble001 , The RIAA feedback network values you wrote must need an edit, because they do not represent an RIAA network.
Having restored and retrofitted my HF-81 and HF-85 phono stages, I recommend this network as an upgrade to the original values, which are close to ideal: 2700pf//2.2Meg. Ohms & 820pf//91K Ohms. This has been objectively verified to be very accurate.

It is simple to measure R and C values with VOM and LCR meters, respectively, in order to match the parts per channel. However, retaining these values after soldering requires extensive heat sinking methods. I use very small, copper, flat bladed alligator clamps, attached to the part to be soldered, as well as attaching them to any adjacent, touching parts, to keep the precision EQ values intact.

I believe your partial schematic is from an Eico manual. I think your chosen insertion points are correct, but your labeling of start/send and return are flipped. The feedback "starts" by the .025uf cap and returns to the V2 grid. I will dig out my factory scheme and check again...

For Eico lurkers, the phono EQ circuitry on HF-85 and HF-81 Eicos is identical. The HF-81 has one stereo volume control and one balance control, with no Loudness control. The HF-85 preamp uses two volume controls, one per channel, concentrically mounted, with taps for a switchable Loudness control.

Unfortunately, the HF-85 Loudness circuitry was published in error and does not sound right in stock form. However, as a saving grace for me, these loudness taps are a great placement for a scratch filter to use with noisy discs and sources. Since the Eico volume control is directly after the phono stage, before any other inputs, it was logical to implement a high-cut, scratch filter with little to no phase shift of the signal. Usually, scratch filters, and tone controls are placed after further gain stages, causing possible phase shifts at certain frequencies, thereby adding musical coloration changes...

AudioKarma does have a superb "Eico Place" thread section, plus the Turntables section is reserved for "All Things Analog" including preamplifiers and phono stages....You might want to check these other AK sections and use the easy Search function to gather more preamp info...

Welcome to AK @drbubble001 , The RIAA feedback network values you wrote must need an edit, because they do not represent an RIAA network.
Having restored and retrofitted my HF-81 and HF-85 phono stages, I recommend this network as an upgrade to the original values, which are close to ideal: 2700pf//2.2Meg. Ohms & 820pf//91K Ohms. This has been objectively verified to be very accurate.

It is simple to measure R and C values with VOM and LCR meters, respectively, in order to match the parts per channel. However, retaining these values after soldering requires extensive heat sinking methods. I use very small, copper, flat bladed alligator clamps, attached to the part to be soldered, as well as attaching them to any adjacent, touching parts, to keep the precision EQ values intact.

I believe your partial schematic is from an Eico manual. I think your chosen insertion points are correct, but your labeling of start/send and return are flipped. The feedback "starts" by the .025uf cap and returns to the V2 grid. I will dig out my factory scheme and check again...

For Eico lurkers, the phono EQ circuitry on HF-85 and HF-81 Eicos is identical. The HF-81 has one stereo volume control and one balance control, with no Loudness control. The HF-85 preamp uses two volume controls, one per channel, concentrically mounted, with taps for a switchable Loudness control.

Unfortunately, the HF-85 Loudness circuitry was published in error and does not sound right in stock form. However, as a saving grace for me, these loudness taps are a great placement for a scratch filter to use with noisy discs and sources. Since the Eico volume control is directly after the phono stage, before any other inputs, it was logical to implement a high-cut, scratch filter with little to no phase shift of the signal. Usually, scratch filters, and tone controls are placed after further gain stages, causing possible phase shifts at certain frequencies, thereby adding musical coloration changes...

AudioKarma does have a superb "Eico Place" thread section, plus the Turntables section is reserved for "All Things Analog" including preamplifiers and phono stages....You might want to check these other AK sections and use the easy Search function to gather more preamp info...

Thanks for an extremely informative and gracious response. I will definitely look up suggested 'Eico Place' thread section. Still have much to learn. Appreciate handy heat sinking tip vis a vis constructing and preserving integrity of RIAA network components.

Schematic is from the HF-81 manual and crude markup of insert points was completely dyslexic error on my part.

 

[tubeactive]

@drbubble001 , You are very welcome !...Enjoy your HF-81 resto...

Here is a link to an Eico adventure, a tale of two Eicos, the HF-85 and HF-81:
https://www.lencoheaven.net/forum/index.php?topic=25066.0

The Eico retro-resto starts after some pre-RIAA phono playback curve discussion. I still regularly use those HF-85 and HF-81 Eicos...

 

[drbubble001]

@tubeactive, thanks for link to your discussion of RIAA etc. it clarified a lot of lingering questions I had. especially useful was turnover calculation 'made simple'. tried to get through Stanley Lipshitz essay yesterday, but got a bit lost. After reading your treatise, I'll try to revisit.

was also unclear re lo freq time constant on Eico phono eq network. which by my calc, yields about 26hz, roughly half of RIAA 50hz spec. calculations using 2700pf//2.2Meg RC, t=5940u freq=26.8hz and turnover @ 517hz. High freq 820pf//91K time constant practically spot on @ 74.62u, 2132hz.

I understand that, as @dcgillespie pointed out, it is unlikely to achieve perfect spec given all the variables in any particular system, especially unpredictable phase anomalies inherent in any eq system. a lot has to do with tweaking to ones taste to achieve desired outcome as you aptly noted.

btw; restoration of your Eico, truly inspiring. if function follows form, it must be an amazing sounding kit.

 

[tubeactive]

Thank YOU @drbubble001 ! Your appreciation of the phono EQ essay is very thoughtful. Your response is a large reason why I write about my audio adventures. If I can inspire others and help them understand more about HiFi, I feel successful. We must share the knowledge...

The evolution of phono EQ curves is quite involved. As LPs and 45s arrived in the late 1940s, there were no "standards" for all record companies to adhere with. There were a few choice guidelines, but the home hifi industry seemed almost surprised when the record "speed wars" began.

Some of the first phono stages for magnetic cartridges utilized very high value, grid resistors to achieve tube bias, with the cathodes grounded. This required an input capacitor as well. With this topology, full frequency response was a designer's chore. Low bass was difficult to provide. Some circuit designers over-accentuated the bass networks, in order to provide flat frequency response results.

Flash forward to 1953, when the AES recommended the RCA New Orthophonic phono EQ as the industry standard. In 1954, the RIAA concurred and RCA's New Ortho EQ became our standard.

Circuitry evolved quickly in the early 1950s. With the New Ortho / RIAA EQ "resonating" at 50Hz, many manufacturers realized if they designed the bass EQ networks to resonate even lower in frequency, bass response measured better under test lab conditions.

When stereo discs really commenced, in 1958, turntable rumble became a new problem to conquer. Noiseless mechanicals and bearings were essential. Eico HF-81 and HF-85 were among the first stereo amps and preamps. Eico engineers chose the lower bass resonance wisely, as they also designed in a rumble filter, using low value coupling caps to minimize any rumbly noises encountered with disc playback.

Researching playback curve design during the early '50s, you will find that many companies had also chosen to design phono networks with lower than 50Hz "resonance."

Designing and simulating circuits is one aspect of implementation. Testing results of finished units either verify proper design or alert the engineers their designs need re-engineering...

One of the earliest phono playback EQ curves, the NAB curve, synonymous with the NBC Orthacoustic curve for those 16 inch transcription discs, brought us the term "shelving" regarding low bass response. With bass boost required, low bass should "level off." If not, rumbly and mechanical noise could become problematic. This later NAB curve evolved into the Columbia LP EQ curve. The Col. LP EQ specified 100Hz as the shelving point. The NAB curve specified 67Hz as the shelving point.

RCA's New Orthophonic curve, later simply called the RIAA curve, specifies 50Hz as the shelving point. Since designing in a linear, low bass "shelf" would necessitate some seriously complex circuitry, engineers chose to "resonate" the low bass EQ at lower than needed frequencies. As mentioned, test results influenced design elements.

Some engineers would refer to the 27Hz resonance as the -3db design point of the Eico bass boost network. Hopefully, this helps explain the reasoning designers were faced with...