MGA-13516-TR1G: GaAs LNA Specs, Replacements & Design Guide

The GaAs LNA That Cellular Infrastructure Kept Running Long After Broadcom Discontinued It

Broadcom listed the MGA-13516 family as obsolete. DigiKey marks it as "Obsolete item." The 2018 EEVblog thread titled "Broadcom/Avago GaAs MMICs now all NRND or obsolete" documented the end of the entire product line. By every official measure, the MGA-13516-TR1G is a discontinued component.

And yet search traffic for MGA-13516-TR1G remains steady years after discontinuation. Secondary market brokers list inventory. RF engineers post questions about it on forums. The explanation is straightforward: the MGA-13516 went into hundreds of cellular base station receiver designs, repeaters, and distributed antenna systems between roughly 2005 and 2015. Those systems are still in the field. When a PCB fails in a deployed BTS cabinet, the repair technician needs the part that is on the board — not the part Broadcom recommends for new designs. A GaAs MMIC with validated performance in a specific gain stage cannot simply be swapped for a different device without reverification of the gain, noise figure, and linearity of the entire receive chain.

This guide covers what MGA-13516-TR1G is, what it actually delivers at 900 MHz and 1900 MHz, why it was designed the way it was, what to watch for in secondary market procurement, and — critically — what the validated migration path looks like for designs that need to move forward.

1.0 What the MGA-13516-TR1G Is

Full designation: Broadcom (formerly Avago Technologies, formerly Agilent Technologies Semiconductor Products) MGA-13516, tape-and-reel packaging, RoHS-compliant, lead-free.

Part number decode:

Field	Characters	Meaning
Family prefix	MGA	Microwave/GaAs Amplifier — Avago's GaAs RFIC product prefix
Product	135	Device family (135xx = 400MHz–1.5GHz high-linearity LNA family)
Variant	16	Specific device within the 135xx family; 16 = 16-pin QFN package version
Packaging	-TR1G	TR = Tape & Reel; 1 = 1,000 units per reel; G = RoHS compliant, lead-free (Pb-free) finish

Companion packaging variants:

• MGA-13516-BLKG = Bulk (tray) packaging, same RoHS/lead-free specification
• MGA-13516-TR2G = Tape & Reel, 3,000 units per reel

All three variants are the same silicon die with identical electrical performance. The suffix encodes only packaging and reel size.

Family context:

The MGA-13516 sits within Avago's MGA-135xx LNA family optimized for cellular infrastructure (400 MHz to 1.5 GHz). The sibling MGA-13316 covers a wider band extending to approximately 4 GHz. Both use Avago's 0.25 µm GaAs E-pHEMT (Enhancement-mode pseudomorphic High Electron Mobility Transistor) process — the same process used in the ATF-54143 discrete pHEMT — in a two-stage integrated MMIC configuration. The 16-pin QFN package (versus the 4-pin SOT-343 or 6-pin SOT-363 used for simpler gain block MMICs) accommodates the two-stage architecture's separate supply and bias pins for each amplifier stage.

---

2.0 Specifications

From the Avago MGA-13516 product datasheet (AV02-xxxx series), at typical operating condition VDD1 = VDD2 = 5V:

At 900 MHz (primary cellular band specification point):

• Gain (S21): approximately 26–28 dB typical
• Noise Figure: approximately 0.5–0.7 dB typical
• OIP3 (Output 3rd-order Intercept Point): approximately 37–39 dBm typical
• OP1dB (Output power at 1dB gain compression): approximately 22–24 dBm typical
• Supply current (IDD1 + IDD2): approximately 100–120 mA total (two stages)
• Input return loss (S11): approximately −14 to −18 dB

At 1900 MHz (PCS/WCDMA band):

• Gain: approximately 24–26 dB typical
• Noise Figure: approximately 0.7–1.0 dB typical
• OIP3: approximately 35–37 dBm typical

Absolute maximum ratings:

• VDD: 6V maximum
• IDD: Limited by thermal dissipation of the 16-QFN package (θJA ≈ 35–40°C/W)
• Input RF power: +15 dBm maximum (damage level)
• Operating temperature: −40°C to +85°C junction

Package:

• 16-QFN, 4.0 mm × 4.0 mm × 0.85 mm body
• Exposed thermal pad (must be soldered to ground plane for proper thermal and electrical performance)
• 0.5 mm lead pitch
• MSL1 moisture sensitivity level (no pre-bake required before reflow)

Process: 0.25 µm GaAs Enhancement-mode pHEMT (E-pHEMT)

---

3.0 How It Works: GaAs E-pHEMT and the Two-Stage LNA Architecture

Why GaAs instead of silicon at 400 MHz–1.5 GHz:

GaAs (Gallium Arsenide) has higher electron mobility than silicon — approximately 5–6× higher. Higher electron mobility means carriers move faster through the semiconductor under the same electric field, enabling lower noise figure and higher frequency operation at the same feature size. At cellular frequencies (400 MHz to 2 GHz), GaAs E-pHEMT technology achieves noise figures in the 0.3–0.8 dB range that silicon CMOS cannot match at equivalent power consumption levels. In a base station receive chain where the received signal may be −100 dBm or weaker, a 0.3 dB difference in LNA noise figure directly translates to 0.3 dB of system sensitivity — which is the difference between receiving a call at the cell edge or dropping it.

E-pHEMT vs standard pHEMT:

Standard pHEMT transistors require a negative gate bias voltage to set the quiescent current (depletion-mode devices). This complicates the bias circuit design — a negative supply rail or a gate-to-source resistor divider is needed. Enhancement-mode pHEMT (E-pHEMT) devices are normally OFF and require a small positive gate voltage (typically 0.4–0.6V) to turn on, allowing operation from a single positive supply. The MGA-13516's internal bias circuit generates the appropriate gate voltages from the positive VDD supply, making it a "single-supply" device despite the two-stage architecture.

Two-stage architecture:

The MGA-13516 integrates two amplifier stages on a single die. The first stage is optimized for minimum noise figure — input matching and bias are set for best NF at the expense of some gain compression headroom. The second stage provides additional gain and determines the linearity ceiling (OIP3, OP1dB). Splitting the supply into VDD1 (first stage) and VDD2 (second stage) allows independent optimization: VDD2 can be set higher than VDD1 to bias the output stage for maximum linearity, or one stage can be powered down independently for power saving.

---

4.0 External Circuit Requirements: Bias, Matching, and Decoupling

Supply decoupling (critical for stability):

Each VDD pin (VDD1 and VDD2) requires both bulk and high-frequency bypass capacitance placed as close as physically possible to the package:

• Bulk bypass: 100 nF X5R/X7R ceramic per VDD pin
• High-frequency bypass: 10–22 pF C0G/NP0 ceramic per VDD pin, placed within 1 mm of the pad

Without adequate high-frequency bypass, the supply line's parasitic inductance creates an impedance at RF frequencies that can cause gain peaking, oscillation, or significantly degrade noise figure. This is the most common cause of MGA-13516 instability in field designs.

Input matching:

The MGA-13516 is specified as requiring "minimum matching" at its input — the internal matching network brings the input impedance close to 50 Ω without external components in many applications. For optimum noise figure (which requires noise-matching, not conjugate matching), a small series inductance at the input can improve NF by 0.1–0.3 dB. Avago's demonstration board schematic (included in the datasheet) shows the specific matching component values for 900 MHz and 1900 MHz.

Output matching:

The output is internally matched to approximately 50 Ω. AC coupling capacitors (typically 100 pF C0G at the input and output) provide DC blocking.

Bias enable/shutdown:

Unlike some gain blocks that operate continuously from a single supply, the MGA-13516's two-stage architecture allows power reduction by disabling VDD2 when the second stage's linearity margin is not needed. This is particularly useful in time-division systems (TDD) where the receive chain can reduce power during transmit slots.

---

5.0 ⚠️ Four Pitfalls — in Both Legacy Maintenance and Replacement Designs

Pitfall 1: Substituting MGA-13316 for MGA-13516 assuming they are interchangeable

The MGA-13316 and MGA-13516 share the same 16-QFN package and are from the same product family, but they are not drop-in substitutes. The MGA-13316 has a wider frequency range (extending to ~4 GHz vs MGA-13516's ~1.5 GHz) achieved through different internal matching networks. The input and output impedance vs frequency profiles differ, meaning that at 900 MHz, the MGA-13316 may not achieve the same noise figure or gain as the MGA-13516 in the same PCB layout with the same matching components. Before substituting 13316 for 13516 (or vice versa) in an existing design, verify gain, NF, and return loss at the operating frequency with the actual PCB matching network.

Pitfall 2: Using secondary-market MGA-13516 parts without verification

Secondary market MGA-13516 inventory carries significant counterfeit risk. The 16-QFN package is straightforward to relabel, and lower-performance devices (including functional GaAs MMICs from other vendors with lower gain or worse NF) can be remarked. In a gain stage where the expected NF is 0.6 dB, a counterfeit part delivering 2.0 dB NF can go undetected through visual inspection and even casual functional testing if the system engineer does not specifically measure noise figure on every received lot. For maintenance repair of installed BTS equipment, test received parts against the datasheet specifications — at minimum, measure gain and P1dB with a VNA and signal source before installing on a production board.

Pitfall 3: Inadequate exposed-pad soldering causing thermal and RF ground issues

The 16-QFN's exposed thermal pad is electrically connected to the device ground and is also the primary thermal path to the PCB. If the exposed pad is not properly reflowed to the PCB ground plane (due to insufficient solder paste, a flat, non-compliant PCB pad, or inadequate reflow profile), two problems occur simultaneously: the device runs hotter than specified (increasing noise figure and potentially causing early failure), and the RF ground reference is poorly defined (increasing gain variation and potentially causing oscillation). Verify exposed-pad solder joints using X-ray inspection on new PCB assemblies.

Pitfall 4: Designing the replacement LNA without re-measuring the system noise figure

When migrating from MGA-13516 to a substitute device (such as the MGA-631P8 described in Section 6), the new device may have slightly different gain, NF, and S-parameters than the original even if specifications appear similar on paper. The system noise figure is set by the cascade: LNA NF contribution + (subsequent stage NF − 1) / LNA Gain (Friis formula). Even a 0.1 dB change in LNA gain can shift the system NF by a small but potentially specification-relevant amount. After replacing the LNA, re-measure the system receiver noise figure before returning the equipment to service.

---

6.0 Modern Substitutes and the Validated Migration Path

When Broadcom discontinued the MGA-135xx family, they published an End-of-Life replacement guide that directly maps discontinued parts to recommended successors. For the MGA-13516 (400 MHz–1.5 GHz high-linearity LNA), the primary validated replacement is:

MGA-631P8 (Broadcom, active production):

• Frequency range: 400 MHz to 1.5 GHz — exact match
• Gain at 900 MHz: 17.5 dB (lower than MGA-13516's ~27 dB — see note below)
• Noise Figure at 900 MHz: 0.53 dB — slightly better
• OIP3 at 900 MHz: 32.6 dBm
• Package: 2×2 mm 8-lead package (smaller than 16-QFN)
• Supply: 4V, 54 mA (lower power than MGA-13516's ~100 mA)

The gain difference is the critical design consideration: The MGA-13516 delivers approximately 26–28 dB gain; the MGA-631P8 delivers approximately 17.5 dB. This is not a pin-compatible replacement — it requires redesigning the gain stage and potentially adding a second gain stage to recover the missing ~10 dB. The NF and linearity improvements partially compensate, but the gain budget must be explicitly re-calculated for the system.

Alternative replacements for evaluation:

Part	Source	Gain @900MHz	NF @900MHz	OIP3 @900MHz	Package
MGA-631P8	Broadcom	17.5 dB	0.53 dB	32.6 dBm	2×2mm 8L
MGA-13116	Broadcom (NRND)	~26 dB	~0.5 dB	~37 dBm	16-QFN
PMA2-162LN+	Mini-Circuits	17.5 dB	0.5 dB	~32 dBm	2×2mm
QPL9547	Qorvo	20 dB	0.6 dB	34 dBm	2×2mm QFN

MGA-13116 caveat: This is listed as NRND (Not Recommended for New Designs) rather than fully obsolete — making it a closer functional match but still a device approaching end-of-life.

For legacy maintenance where footprint change is not feasible, sourcing authentic MGA-13516 from verified secondary market brokers with test reports is the most pragmatic short-term solution. For new or redesigned boards, the MGA-631P8 (with gain stage redesign) represents the intended migration path.

---

7.0 PCB Design Notes for the 16-QFN Package

Footprint: The MGA-13516's 16-QFN (4×4 mm) footprint follows JEDEC MO-220 outline with 16 leads at 0.5 mm pitch and a center exposed thermal pad. Use Avago's recommended land pattern from the package specification document.

RF trace routing: Input and output RF traces should be 50 Ω microstrip width for the PCB stackup in use (typically 2.5–3.0 mm wide on 1.6 mm FR4 for 50 Ω, or 1.0–1.5 mm on 0.8 mm Rogers RO4350B). Keep RF input and output traces short, on the same layer, with no vias in the signal path between the package pad and the first RF component (DC block capacitor or matching element).

Ground vias: Place multiple ground vias directly adjacent to the exposed thermal pad connection to the ground plane, and also adjacent to each ground pad on the package perimeter. Inadequate ground via density introduces parasitic inductance at the ground connection, which shifts the gain flatness and can cause low-frequency oscillation.

Layer stackup: For best performance at 1.5 GHz, use at minimum a 4-layer PCB with the RF signal layer directly above an unbroken ground plane. Rogers RO4350B (εr = 3.66) provides more predictable microstrip impedance than FR4 (εr = 4.3–4.5, variable) and is appropriate for production cellular infrastructure PCBs.

---

8.0 Sourcing MGA-13516-TR1G in 2025

Authorized channels: Broadcom no longer manufactures MGA-13516. DigiKey and Mouser list it as obsolete with no stock. Arrow and Avnet list it as end-of-life with no active inventory.

Secondary market: Independent component distributors and broker networks (IHS Markit/Supplyframe, PartMiner, 4Star Electronics, Converge) may carry MGA-13516-TR1G from excess stock from original manufacturers' runs or from OEM/EMS inventory liquidation. Pricing in secondary market at obsolete status typically runs 3–8× the original production price per unit.

Incoming inspection recommendations for secondary market parts:

• Visual: examine package markings under 40× magnification; compare date codes to known production windows
• Electrical: measure S21 (gain) at 900 MHz — should be approximately 25–28 dB at VDD = 5V, IDD ~110 mA
• Noise figure measurement: NF should be below 1.0 dB at 900 MHz; values above 1.5 dB suggest counterfeit or degraded parts
• P1dB verification: output 1dB compression should be above 20 dBm

For verified MGA-13516-TR1G with traceability documentation and incoming test records, visit aichiplink.com.

---

9.0 Real Questions from RF Engineers

Q: The receive chain sensitivity in our BTS cabinet has degraded by approximately 1.5 dB. The LNA is an MGA-13516. We replaced it with a new MGA-13516 from secondary market and the sensitivity is still degraded. What else should we check?

A: A 1.5 dB system sensitivity degradation with a functioning LNA replacement points to several possibilities in order of likelihood. First, check the LNA's supply voltage under operating conditions — a marginal power supply or excessive supply line resistance can reduce VDD below 5V, which increases noise figure by 0.3–0.8 dB at 4.5V versus 5.0V. Second, check the input path for damaged components: SAW filter degradation or poor connector contact before the LNA adds directly to system noise figure. Third, verify the secondary market replacement is actually performing to spec — measure gain (should be ~26 dB) and if possible noise figure with a noise figure meter or Y-factor method. A genuine MGA-13516 with adequate decoupling and proper bias should restore sensitivity. If it does not, the degradation is upstream of the LNA.

Q: We have 500 MGA-13516-TR1G units from a broker that we purchased for service stock. How do we verify they are authentic before we need them in the field?

A: Test a sample of 10–20 units (minimum 2% of lot) at operating conditions. Measure: (1) DC supply current at VDD1 = VDD2 = 5V — should be 90–130 mA total; (2) S21 gain at 900 MHz using a VNA — should be 25–28 dB; (3) P1dB at 900 MHz — should be above 20 dBm. Parts failing any of these by more than 2 dB should be rejected as counterfeit or degraded. Document the test results and retain with the lot for traceability. If the full lot passes incoming inspection, store in ESD-safe packaging at controlled humidity (MGA-13516 is MSL1 — no moisture exposure limitation, but cleanliness and ESD protection are important).

Q: Is there any active-production GaAs MMIC that is a true pin-compatible drop-in for MGA-13516 in the same 16-QFN footprint?

A: Not from Broadcom's current portfolio. The MGA-631P8 (the recommended Broadcom successor) uses a 2×2 mm 8-lead package — not pin-compatible with the 16-QFN. The MGA-13116 (NRND, same 16-QFN) is the closest functional and footprint match, but it is itself approaching end-of-life. No third-party vendor currently produces a 16-QFN pin-compatible replacement with equivalent specifications. For designs where the 16-QFN footprint must be maintained on an existing PCB (service board replacement or PCB revision not feasible), secondary market MGA-13516 procurement with incoming inspection is the most practical path. For any new PCB revision, migrate to the 2×2 mm package generation — the smaller package provides equal or better RF performance and is supported by active production parts from Broadcom and alternatives.

---

10.0 Quick Reference Card

Part Number Decode:

Field	Value	Meaning
MGA	MGA	Microwave/GaAs Amplifier prefix
135	135	400MHz–1.5GHz LNA family
16	16	16-pin QFN package variant
TR	TR	Tape and Reel
1	1	1,000 units per reel
G	G	RoHS compliant, lead-free

Key Specifications (typical, VDD = 5V, 900 MHz):

Parameter	Typical Value
Frequency range	400 MHz – 1.5 GHz
Gain (S21)	~27 dB
Noise Figure	~0.6 dB
OIP3	~38 dBm
OP1dB	~22 dBm
Supply voltage	5V (4.5–5.5V)
Total IDD	~110 mA (VDD1 + VDD2)
Package	16-QFN, 4×4 mm
Status	Obsolete (DigiKey, Mouser, 2018+)

Replacement Options:

Part	Status	Gain	NF	Package	Notes
MGA-631P8	Active	17.5 dB	0.53 dB	2×2mm 8L	Broadcom validated replacement; gain stage redesign needed
MGA-13116	NRND	~26 dB	~0.5 dB	16-QFN	Same footprint; approaching EOL
PMA2-162LN+	Active	17.5 dB	0.5 dB	2×2mm	Mini-Circuits alternative
QPL9547	Active	20 dB	0.6 dB	2×2mm QFN	Qorvo; closer gain match

Incoming Inspection Checklist (secondary market):

• IDD at VDD1 = VDD2 = 5V: 90–130 mA
• S21 at 900 MHz: 25–28 dB
• P1dB at 900 MHz: > 20 dBm
• NF at 900 MHz: < 1.0 dB (reject if > 1.5 dB)

---

For sourcing Broadcom MGA-13516-TR1G with traceability documentation and incoming test records, visit aichiplink.com.

Search Broadcom MGA-13516-TR1G Stock Now

 

 

 

 

 

Written by Jack Elliott from AIChipLink.

 

AIChipLink, one of the fastest-growing global independent electronic   components distributors in the world, offers millions of products from thousands of manufacturers, and many of our in-stock parts is available to ship same day.

 

We mainly source and distribute integrated circuit (IC) products of brands such as Broadcom, Microchip, Texas Instruments, Infineon, NXP, Analog Devices, Qualcomm, Intel, etc., which are widely used in communication & network, telecom, industrial control, new energy and automotive electronics. 

 

Empowered by AI, Linked to the Future. Get started on AIChipLink and submit your RFQ online today!